{"schemaVersion":2,"updatedAt":"2026-09-23T20:02:39.062013Z","items":[{"id":"sw-3822e53281df1dc0922d","title":"Deprecation notice: All-platform CodeQL bundle","summary":"Starting with CodeQL CLI 2.27.0, the all-platform CodeQL bundle (i.e., codeql-bundle.tar.gz and codeql-bundle.tar.zst), which includes the binaries for all supported platforms up to this release, is marked as deprecated. In… The post Deprecation notice: All-platform CodeQL bundle appeared first on The GitHub Blog.","url":"https://github.blog/changelog/2026-09-22-deprecation-notice-all-platform-codeql-bundle","source":"GitHub Changelog","publishedAt":"2026-09-22T09:21:03Z","category":"developer","tags":[],"importance":"mustKnow","clusterCount":1},{"id":"sw-975d40417c9e98826fb3","title":"CISA Adds One Known Exploited Vulnerability to Catalog","summary":"CISA has added one new vulnerability to its Known Exploited Vulnerabilities (KEV) Catalog, based on evidence of active exploitation. CVE-2026-7273 Zyxel GS1900 Series Switches Stack-Based Buffer Overflow Vulnerability This type of vulnerability is a frequent attack vector for malicious cyber actors and poses significan","url":"https://cisa.gov/news-events/alerts/2026/09/21/cisa-adds-one-known-exploited-vulnerability-catalog","source":"CISA Advisories","publishedAt":"2026-09-21T12:00:00Z","category":"security","tags":[],"importance":"mustKnow","clusterCount":1},{"id":"sw-aeff1dc7e91a24bb2d1c","title":"Windows Exploitation Techniques: Dangling COM Object Registrations","summary":"This short blog post is about abusing a privilege escalation bug that Microsoft recently fixed in Windows, CVE-2026-66804, that I and 14 others reported. This issue is an incomplete fix for CVE-2026-50343, a bug dubbed “Dark Elevator” by Calif. The root cause of the bug was a dangling COM object registration for the Cr","url":"https://projectzero.google/2026/09/windows-dangling-com.html","source":"Google Project Zero","publishedAt":"2026-09-21T07:00:00Z","category":"security","tags":[],"importance":"mustKnow","clusterCount":1},{"id":"sw-849d1c74e3ee1c32cb80","title":"SpeakerMem-R1: Speaker-Centered Dual-Track Memory for Multi-Party Dialogue","summary":"Long-term conversational memory in multi-party settings requires more than retrieving relevant content from long-term conversations: it must distinguish who said what, whom each statement concerns, how individuals perceive one another, what information is shared by the group, and how states change over time. Recent stu","url":"https://arxiv.org/abs/2609.26780v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T17:56:12Z","category":"research","tags":[],"importance":"important","clusterCount":3},{"id":"sw-c99106a5c97c0f2896ac","title":"Gemini 3.8 text-to-speech","summary":"","url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-3-8-text-to-speech","source":"Hacker News","publishedAt":"2026-09-23T15:29:23Z","category":"developer","tags":[],"importance":"important","clusterCount":1},{"id":"sw-b7bf46131e8f925848fa","title":"TransBERT: A Framework for Synthetic Translation in Domain-Specific Language Modeling","summary":"The scarcity of non-English language data in specialized domains significantly limits the development of effective Natural Language Processing (NLP) tools. We present TransBERT, a novel framework for pre-training language models using exclusively synthetically translated text, and introduce TransCorpus, a scalable tran","url":"https://arxiv.org/abs/2609.26347v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T12:53:35Z","category":"research","tags":[],"importance":"important","clusterCount":3},{"id":"sw-282fc76ce2e343d2b5c6","title":"CISA Adds Four Known Exploited Vulnerabilities to Catalog","summary":"CISA has added four new vulnerabilities to its Known Exploited Vulnerabilities (KEV) Catalog, based on evidence of active exploitation. CVE-2026-85102 Check Point Multiple Products Improper Certificate Validation Vulnerability CVE-2026-93616 Check Point Multiple Products Path Traversal Vulnerability CVE-2026-93952 Aris","url":"https://cisa.gov/news-events/alerts/2026/09/22/cisa-adds-four-known-exploited-vulnerabilities-catalog","source":"CISA Advisories","publishedAt":"2026-09-22T12:00:00Z","category":"security","tags":[],"importance":"important","clusterCount":1},{"id":"sw-4182359e79e632d3734e","title":"Claude Platform release notes — September 22, 2026","summary":"We've launched Claude Opus 5.5 ( claude-opus-5-5 ), a model for long-running agentic coding and knowledge work. It has a 1M token context window by default, 128k max output tokens, and always-on adaptive thinking, at $4 / $20 USD per MTok (Claude Opus 5 is $5 / $25). Claude Opus 5.5 is available on the Claude API, Clau","url":"https://platform.claude.com/docs/en/release-notes/overview","source":"Claude Release Notes","publishedAt":"2026-09-22T00:00:00Z","category":"products","tags":[],"importance":"important","clusterCount":2},{"id":"sw-de0493e40ab04f045700","title":"Pentagon says overreliance on AI contributed to missile strike on Iran school","summary":"https://archive.ph/0V37g","url":"https://bloomberg.com/graphics/2026-iran-school-attack","source":"Hacker News","publishedAt":"2026-09-22T19:03:38Z","category":"developer","tags":[],"importance":"important","clusterCount":1},{"id":"sw-269ae7db0aca5a22cd2d","title":"CISA Adds Two Known Exploited Vulnerabilities to Catalog","summary":"CISA has added two new vulnerabilities to its Known Exploited Vulnerabilities (KEV) Catalog, based on evidence of active exploitation. CVE-2025-39964 Linux Kernel Race Condition Vulnerability CVE-2026-53266 Linux Kernel Out-of-Bounds Write Vulnerability These types of vulnerabilities are a frequent attack vector for ma","url":"https://cisa.gov/news-events/alerts/2026/09/18/cisa-adds-two-known-exploited-vulnerabilities-catalog","source":"CISA Advisories","publishedAt":"2026-09-18T12:00:00Z","category":"security","tags":[],"importance":"important","clusterCount":1},{"id":"sw-6eff16a1336643fb3f35","title":"v2.1.281","summary":"What's changed Added Claude apps gateway support for newer Claude Desktop keys in desktop policy blocks, including blockReadsOutsideWorkingDirectories and disableBypassPermissionsMode Added assume_role on Claude apps gateway Bedrock upstreams: the gateway calls Bedrock as an IAM role it assumes through STS, in another ","url":"https://github.com/anthropics/claude-code/releases/tag/v2.1.281","source":"Claude Code Releases","publishedAt":"2026-09-23T19:19:15Z","category":"products","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-ef00200e6cca5a2adca2","title":"Google Beam expands with new regions, partners, and customers","summary":"","url":"https://blog.google/innovation-and-ai/technology/research/google-beam-expansion","source":"Google AI","publishedAt":"2026-09-23T18:00:00Z","category":"labs","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-52e46e8371e1f69996c0","title":"Two years of OpenAI Academy","summary":"Marking two years of OpenAI Academy and bringing AI skills to even more communities.","url":"https://openai.com/index/two-years-of-openai-academy","source":"OpenAI News","publishedAt":"2026-09-23T16:00:00Z","category":"labs","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-1043d9fe80283e6f409b","title":"Advancing Private AI Compute with secure, server-side memory","summary":"Introducing private, server-side memory to Private AI Compute for personal AI.","url":"https://deepmind.google/blog/advancing-private-ai-compute-with-secure-server-side-memory","source":"Google DeepMind","publishedAt":"2026-09-23T16:00:57Z","category":"labs","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-45c15a5695991642d974","title":"AI leaders warn UN of security risks as systems grow more powerful - Reuters","summary":"AI leaders warn UN of security risks as systems grow more powerful Reuters","url":"https://news.google.com/rss/articles/CBMivgFBVV95cUxOazV6Q242LXE0TVZZRV9wcXl4UmJlMHp2ZHMtX0h4RGlPeGlUbkhGQklJTDhuN0FKaFdzUUs3cS0zN1ZsTVR1Mk51M190NFFuVjdZekJVSjlXMXpuaDAwVGVaSl9LbnpKMWlmMG1YaUVOQ3lqTXl4ck5nZm94RlMxaFA2N091TUxyNWJCSkpLOER1TDAtRF94TWZ6NVNIM01lcWxvOUJaWTVHYUZ3WlJjSm16WWtUSldqcHdWOFNR?oc=5","source":"Reuters AI","publishedAt":"2026-09-23T19:46:24Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-30aa7c0acfe7f3809f1c","title":"How to Use NVIDIA Warp and MjWarp to Accelerate Robotics Simulation and Learning Workflows","summary":"","url":"https://huggingface.co/blog/nvidia/how-to-use-nvidia-warp-and-mjwarp","source":"Hugging Face","publishedAt":"2026-09-23T18:41:40Z","category":"labs","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-721de6922a1a6b9fd6bd","title":"Gemini 3.8 text-to-speech says hello","summary":"","url":"https://deepmind.google/blog/say-hello-to-gemini-38-text-to-speech","source":"Google DeepMind","publishedAt":"2026-09-23T15:25:14Z","category":"labs","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-9328fa400dd37e6b9e77","title":"OpenAI extends cyber access to Ukraine for civilian defense","summary":"OpenAI is extending access to its Daybreak program to the Government of Ukraine to support the cyber defense of civilian infrastructure.","url":"https://openai.com/index/openai-extends-cyber-access-to-ukraine-for-civilian-defense","source":"OpenAI News","publishedAt":"2026-09-23T13:00:00Z","category":"labs","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-b723d18166247bf58e9e","title":"OECD expects AI boom to help offset Middle East energy shock for now - Reuters","summary":"OECD expects AI boom to help offset Middle East energy shock for now Reuters","url":"https://news.google.com/rss/articles/CBMirgFBVV95cUxNUjhCamFjUEdaSHhrSGlUR3hEaHhoczQtY0d1Q19lNWhsTUdtX0Rxb2FkbjdUSDlmOVlZZ19rdjZ2NDFrT3YyYmt6WG1WdEYtUVBjaWxuVTR1Sk1KYjJxVEppbjZtQ0hfVVpLM1UtV3pqTHJMWkR5ekNLMUczR18tekM4NXJrdHFYOXBCUTdOdjlLVTdjVDZkSThPSm84UXVDMWFLVWZ1OU1oRWRhUGc?oc=5","source":"Reuters AI","publishedAt":"2026-09-23T17:09:57Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-63e0251438a40139e79a","title":"Local sandboxing in the GitHub Copilot app","summary":"Local sandboxing helps reduce the potential impact of unintended commands by limiting access to files, network resources, and credentials on your machine. In the GitHub Copilot app, you configure it… The post Local sandboxing in the GitHub Copilot app appeared first on The GitHub Blog.","url":"https://github.blog/changelog/2026-09-23-local-sandboxing-in-the-github-copilot-app","source":"GitHub Changelog","publishedAt":"2026-09-23T15:00:57Z","category":"developer","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-d25d967d25d3918ac732","title":"CISA Adds One Known Exploited Vulnerability to Catalog","summary":"CISA has added one new vulnerability to its Known Exploited Vulnerabilities (KEV) Catalog, based on evidence of active exploitation. CVE-2025-39682 Linux Kernel Improper Check for Unusual or Exceptional Conditions Vulnerability This type of vulnerability is a frequent attack vector for malicious cyber actors and poses ","url":"https://cisa.gov/news-events/alerts/2026/09/18/cisa-adds-one-known-exploited-vulnerability-catalog","source":"CISA Advisories","publishedAt":"2026-09-18T12:00:00Z","category":"security","tags":[],"importance":"important","clusterCount":1},{"id":"sw-4e56c7312349a760a0ed","title":"Ringg’s AI agents resolve up to 65% of customer calls with OpenAI","summary":"Using GPT-5.6, Ringg powers multilingual agents across voice, chat, WhatsApp, and web for 90% less cost vs. GPT-4.1.","url":"https://openai.com/index/ringg","source":"OpenAI News","publishedAt":"2026-09-23T12:00:00Z","category":"labs","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-52b80abd4e046c041ffb","title":"Harvey turns legal context into stronger drafts with GPT-6 Astra","summary":"GPT-6 Astra produces more structured, context-aware legal documents, freeing lawyers to focus on strategy.","url":"https://openai.com/index/harvey-from-context-to-confidence-with-astra","source":"OpenAI News","publishedAt":"2026-09-23T12:00:00Z","category":"labs","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-e728459e578540e9a88f","title":"Considerations for Critical Infrastructure Operators Working With Third-Party ICS Integrators","summary":"Introduction The Federal Bureau of Investigation (FBI) and Cybersecurity and Infrastructure Security Agency (CISA)—hereafter referred to as the “authoring agencies”—have published this fact sheet to highlight considerations for critical infrastructure entities to reduce risk and minimize vulnerabilities when working wi","url":"https://cisa.gov/resources-tools/resources/considerations-critical-infrastructure-operators-working-third-party-ics-integrators","source":"CISA Advisories","publishedAt":"2026-09-23T12:00:00Z","category":"security","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-f2ff7f76056120414cf0","title":"How invideo improves color grading 3x with GPT‑6 Astra","summary":"With GPT‑6 Astra, invideo plans edits with greater precision, improves color correction and grading threefold, and produces 50 custom effects in one day.","url":"https://openai.com/index/invideo-builds-with-gpt-6-astra","source":"OpenAI News","publishedAt":"2026-09-23T12:00:00Z","category":"labs","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-b2cab94e5375e0ba8f1a","title":"Enveda secures $311M to bring more nature-derived AI drugs into clinical trials","summary":"The round valued the AI biotech at $2 billion. It is currently testing drugs that treat skin conditions and preserve weight loss after stopping GLP-1s.","url":"https://techcrunch.com/2026/09/23/enveda-secures-311m-to-bring-more-nature-derived-ai-drugs-into-clinical-trials","source":"TechCrunch AI","publishedAt":"2026-09-23T19:31:28Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-856b863d6730e131607e","title":"Canadian and French publishers stand by author accused of using AI for prize-winning novel - Reuters","summary":"Canadian and French publishers stand by author accused of using AI for prize-winning novel Reuters","url":"https://news.google.com/rss/articles/CBMivgFBVV95cUxPOHY3NmZmazZOWVpVaHpPMFVHaWUtUWxSUmZhYjNYUEVOZmxuOEgyRl85UWUzMFd4dW1SRTNPQ2E0cUxWemFPcXVmM1lQY1ZCTVF4QnJCeWxRV1hveTZTZmxtMWRtVFE1Zno2enlQbVlNOU9sNUFBV19VSWU5dkFBbVNrOHR1TVJsblRjNVowRjNHa1ctbndpUU1fcFlFM1ZfYjBfX3ljMkJZV2k3UnNIamIxcnpaektPU0xFM2Z3?oc=5","source":"Reuters AI","publishedAt":"2026-09-23T15:18:43Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-3d7b9aab56c8f4fe22c0","title":"Introducing MentalHealthBench","summary":"MentalHealthBench is an expert-informed benchmark for evaluating helpful and safe AI responses across realistic mental health conversations.","url":"https://openai.com/index/introducing-mentalhealthbench","source":"OpenAI News","publishedAt":"2026-09-23T10:00:00Z","category":"labs","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-f4e312471c28cd036c07","title":"Meta’s AI agent is a cute little guy who’s great at spending my money","summary":"Modern life comes with an unending, auto-populating to-do list. It never ceases to amaze me how I can be doing nothing at all, minding my own business, and suddenly something needs to be taken care of. You're telling me a tree branch fell in the backyard and now I have to figure out what to […]","url":"https://theverge.com/ai-artificial-intelligence/999526/meta-muse-ai-agent-hands-on-shopping","source":"The Verge AI","publishedAt":"2026-09-23T18:50:37Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-6300426da5d34dc29b7a","title":"Data centers are black boxes, but California wants to change that","summary":"California Gov. Gavin Newsom signed a slate of bills on Monday that could finally give communities better data - and more say - on how data centers impact their electricity bills and water supply. As data centers invade a growing number of communities across the US, they've triggered protests over how the infrastructur","url":"https://theverge.com/policy/999412/data-center-water-electricity-disclosure-bills","source":"The Verge AI","publishedAt":"2026-09-23T18:22:57Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-d8e11dd54e54d539c184","title":"Amazon rolls out new agentic AI for third-party sellers - Reuters","summary":"Amazon rolls out new agentic AI for third-party sellers Reuters","url":"https://news.google.com/rss/articles/CBMisgFBVV95cUxQNno1MXBaZ3JtY05ZWHpfU2RsV0JuU3Zlb3hlN2RiaExrSHdEbE5wNGZjZUhINUNJQkdBR0pTM0xQUFhHZjdMdmEwSGtOcDRVcjJ1dFJNU1ZqMWlzbFpDZFhUVTlGaDFpTENaRjhKSDk0YWFiX05VOUZNbVVqRGhSemt6MHpabV8xbHNHSkZwbGRWVy1XdFdGZlVUVS1iMEJCNUNkd2stYUU2TnZ6Ny1LSXZB?oc=5","source":"Reuters AI","publishedAt":"2026-09-23T13:03:43Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-4cb1dc983c74e0d66538","title":"ChatGPT mobile app gets voice-based agentic features","summary":"Pro and Plus users will be able to use the Work tab on their phones to complete agentic tasks.","url":"https://techcrunch.com/2026/09/23/chatgpt-mobile-app-gets-voice-based-agentic-features","source":"TechCrunch AI","publishedAt":"2026-09-23T17:00:00Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-51ae79308bb54c33912c","title":"Even Americans who use AI every day are worried about it","summary":"The report suggests that greater exposure will not resolve the unease around the technology, nor reduce public support for AI regulation.","url":"https://techcrunch.com/2026/09/23/even-americans-who-use-ai-every-day-are-worried-about-it","source":"TechCrunch AI","publishedAt":"2026-09-23T16:49:58Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-88a2e4e533d7cb151973","title":"After multiple deaths and injuries, NHTSA is investigating comma.ai","summary":"Feds know of 5 crashes where cars using comma.ai devices hit slow or stopped vehicles.","url":"https://arstechnica.com/cars/2026/09/aftermarket-driver-assist-under-federal-probe-following-fatal-crashes","source":"Ars Technica","publishedAt":"2026-09-23T14:20:23Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-282b125f98d4fe4bab21","title":"Anthropic’s biolab made a discovery it’s comparing to Crispr","summary":"Anthropic says its AI Claude has \"autonomously discovered\" a new enzyme system similar to machinery behind the powerful gene-editing tool Crispr. It's the first result from Anthropic's newly-launched wet lab and an early test of Claude's usefulness for science as the company prepares to go public. The company says Clau","url":"https://theverge.com/ai-artificial-intelligence/999470/anthropic-biolab-claude-crispr","source":"The Verge AI","publishedAt":"2026-09-23T17:10:35Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-975511446d6cdf38449d","title":"Bernie Sanders proposes banning ‘superintelligence’ and putting violators in prison","summary":"Sen. Bernie Sanders (I-VT) and Rep. Greg Casar (D-TX) have introduced new legislation that would ban anyone from developing artificial superintelligence - a technology the bill describes as capable of the \"destruction or disempowerment of humanity,\" including by overthrowing the government. Under the Ban Artificial Sup","url":"https://theverge.com/ai-artificial-intelligence/999443/bernie-sanders-ai-superintelligence-ban-act","source":"The Verge AI","publishedAt":"2026-09-23T16:38:09Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-f665497a7ce06d235867","title":"America gave up its rare earth edge. China took full advantage.","summary":"Decades of cheaper imports left US dependent on China for crucial rare earth processing.","url":"https://arstechnica.com/reviews/2026/09/how-the-us-handed-china-control-of-the-rare-earth-supply-chain","source":"Ars Technica","publishedAt":"2026-09-23T13:37:05Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-5a299724cecafd043f61","title":"YouTube Music gets more conversational with new AI features","summary":"Built directly into the YouTube Music app, Ask Music lets users describe what they want to hear in everyday language rather than searching for individual songs or artists.","url":"https://techcrunch.com/2026/09/23/youtube-music-gets-more-conversational-with-new-ai-features","source":"TechCrunch AI","publishedAt":"2026-09-23T15:19:30Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-baa20b51d4c000edd305","title":"YouTube releases new AI features for creators within its Studio app","summary":"YouTube is adding new features to generate ideas and monitor the performance of thumbnails.","url":"https://techcrunch.com/2026/09/23/youtube-releases-new-ai-features-for-creators-within-its-studio-app","source":"TechCrunch AI","publishedAt":"2026-09-23T14:30:00Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-dedd20cb4e6c17ac04c0","title":"StrictlyVC at TechCrunch Disrupt 2026: Inside the changing rules of venture capital","summary":"StrictlyVC joins TechCrunch Disrupt 2026 to discuss the changing VC landscape thanks to AI. Get your Investor Pass to join these exclusive sessions. Save $200 before September 25 at 11:59 p.m. PT.","url":"https://techcrunch.com/2026/09/23/strictlyvc-at-techcrunch-disrupt-2026-inside-the-changing-rules-of-venture-capital","source":"TechCrunch AI","publishedAt":"2026-09-23T14:30:00Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-e14978ff9b482fd8728f","title":"YouTube will let you build your own algorithm with AI","summary":"YouTube’s new custom feeds let users describe the videos they want to see in their own words, then use Gemini to build a personalized feed around the request.","url":"https://techcrunch.com/2026/09/23/youtube-will-let-you-build-your-own-algorithm-with-ai","source":"TechCrunch AI","publishedAt":"2026-09-23T14:30:00Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-dcce20a8c56d15e1e8ec","title":"3 days left to save up to $200 and make impactful connections at TechCrunch Disrupt 2026","summary":"3 days to save up to $200 on your TechCrunch Disrupt 2026 pass, plus 50% off a second. Make impactful connections with 10,000+ tech leaders. Last day to save is September 25 at 11:59 p.m. PT. Register today.","url":"https://techcrunch.com/2026/09/23/3-days-left-to-save-up-to-200-and-make-impactful-connections-at-techcrunch-disrupt-2026","source":"TechCrunch AI","publishedAt":"2026-09-23T14:00:00Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-c1dd3a74eb826fd88713","title":"OpenAI nabs key Patreon execs ahead of upcoming announcement","summary":"OpenAI has hired three former Patreon execs to anchor its product strategy for creators. After starting the creator subscription platform 13 years ago, cofounder and technology chief Sam Yam announced on X that he's joining OpenAI to lead Creator Product. He's also bringing Patreon's former product head Drew Rowny and ","url":"https://theverge.com/ai-artificial-intelligence/999249/openai-creators-patreon-execs-hire-sam-yam","source":"The Verge AI","publishedAt":"2026-09-23T14:35:12Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-2bccfcbede8845ca6d35","title":"How we made claude.ai 3x faster in two weeks","summary":"","url":"https://claude.dev/blog/how-we-made-claude-ai-faster","source":"Hacker News","publishedAt":"2026-09-23T19:23:40Z","category":"developer","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-f3bb9bf2f30738f115be","title":"Small undersea volcanoes may unleash outsized tsunamis","summary":"Post-eruption imaging of the Hunga caldera gives new details of its collapse.","url":"https://arstechnica.com/science/2026/09/small-undersea-volcanoes-may-unleash-outsized-tsunamis","source":"Ars Technica","publishedAt":"2026-09-23T11:15:30Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-041e3e34045aec6ee1b6","title":"A new report from Europe raises serious alarms about orbital collisions","summary":"\"I think we are going to lose the ability to use certain orbits.\"","url":"https://arstechnica.com/space/2026/09/there-are-now-17000-satellites-in-orbit-and-growing-should-we-be-concerned","source":"Ars Technica","publishedAt":"2026-09-23T11:00:17Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-0d30a54669d6d323759f","title":"Spotify is giving you the keys to its recommendation algorithm with US launch of ‘Taste Profile’","summary":"Spotify is rolling out Taste Profile to Premium users in the U.S., letting listeners see how the streamer understands their tastes and use natural language to reshape their recommendations.","url":"https://techcrunch.com/2026/09/23/spotify-is-giving-you-the-keys-to-its-recommendation-algorithm-with-u-s-launch-of-taste-profile","source":"TechCrunch AI","publishedAt":"2026-09-23T13:00:00Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-f146131f3046d669d65a","title":"The AI Hype Index: AI loves cheating","summary":"Brace yourself: It turns out AI is being optimized for cheating. OpenAI’s agents hacked into Hugging Face to get the answers to a cybersecurity test. Next, they solved a prestigious math problem (or just stole from two top mathematicians’ answer sheets). Anthropic’s models have also hacked into other companies’ systems","url":"https://technologyreview.com/2026/09/23/1144940/ai-hype-index-ai-loves-cheating","source":"MIT Technology Review","publishedAt":"2026-09-23T09:00:00Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-9ccedb3b35ab1340d1e3","title":"YouTube is building AI creator tools that do almost everything for them","summary":"Part of the job of a content creator is to figure out how to get their work in front of the most people. Cracking - or fighting - the algorithm has historically been a frustration for creators, but YouTube is increasingly simply telling creators what they should do. At the annual creator-focused Made on YouTube […]","url":"https://theverge.com/tech/999140/made-on-youtube-creator-tools-ai-thumbnails-tests","source":"The Verge AI","publishedAt":"2026-09-23T13:49:24Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-d8ad8b88550b2dd75dcc","title":"Cloud Agents Are Inevitable AI Prisons","summary":"","url":"https://normanponte.io/19df691f","source":"Hacker News","publishedAt":"2026-09-23T18:16:41Z","category":"developer","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-1898b8cd3bcc24490613","title":"Ema raises $77M as AI starts eating into enterprise software and services","summary":"Ema has raised $140 million to date and has more than 50 enterprise customers, including Google and Microsoft.","url":"https://techcrunch.com/2026/09/23/ema-raises-77m-as-ai-starts-eating-into-enterprise-software-and-services","source":"TechCrunch AI","publishedAt":"2026-09-23T12:00:00Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-0c84c3c9918ad90b7285","title":"Claude discovers a novel enzyme system with CRISPR-like repeats","summary":"","url":"https://anthropic.com/news/claude-discovers-novel-enzyme-system","source":"Anthropic News","publishedAt":"2026-09-23T00:00:00Z","category":"labs","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-5ab590850491008e2cef","title":"Grab and OpenAI bring practical AI skills to Southeast Asia","summary":"OpenAI and Grab launch GO Forward with AI, a regional programme helping 30,000 partners build practical AI skills across Southeast Asia.","url":"https://openai.com/index/grab-openai-ai-skills-southeast-asia","source":"OpenAI News","publishedAt":"2026-09-23T00:00:00Z","category":"labs","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-ef7fe9a5febaf0691733","title":"Community condemn Bedminster AI mural as 'insult' to artists","summary":"","url":"https://bbc.co.uk/news/articles/cmlyq09jp49eo","source":"Hacker News","publishedAt":"2026-09-23T14:16:12Z","category":"developer","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-cf4f876c079667faff55","title":"OpenTelemetry in the GitHub Copilot app","summary":"Understand how Copilot agents perform and interact with models and tools. The GitHub Copilot app now supports OpenTelemetry (OTel) configuration through enterprise-managed settings. OTel is an open source observability framework.… The post OpenTelemetry in the GitHub Copilot app appeared first on The GitHub Blog.","url":"https://github.blog/changelog/2026-09-22-opentelemetry-in-the-github-copilot-app","source":"GitHub Changelog","publishedAt":"2026-09-23T02:14:54Z","category":"developer","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-2e8e9ffd298aec64b113","title":"Jev in practice: typed decisions, scoped authority","summary":"","url":"https://tenuo.ai/blog/jev-scoped-authority","source":"Hacker News","publishedAt":"2026-09-23T14:11:14Z","category":"developer","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-69839069a6e7ae540727","title":"Stripe's Knowledge AI Platform","summary":"","url":"https://stripe.dev/blog/meet-stripes-knowledge-ai-platform","source":"Hacker News","publishedAt":"2026-09-23T13:38:00Z","category":"developer","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-cc47574a679086e34b15","title":"What AI-Native Looks Like","summary":"","url":"https://fagnerbrack.com/heres-what-ai-native-with-stable-infra-looks-like-9f17cdb61459","source":"Hacker News","publishedAt":"2026-09-23T12:54:38Z","category":"developer","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-18036a2d23f1f8494acf","title":"New features and improvements in Copilot for JetBrains","summary":"GitHub Copilot for JetBrains 1.18.0 brings AI-assisted tool approvals, more control over agent conversations, and shared skills and instructions for your organization. You can also review plans with the Codex… The post New features and improvements in Copilot for JetBrains appeared first on The GitHub Blog.","url":"https://github.blog/changelog/2026-09-22-new-features-and-improvements-in-copilot-for-jetbrains","source":"GitHub Changelog","publishedAt":"2026-09-23T00:34:18Z","category":"developer","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-d2818a26d65a75a81ed6","title":"OpenAI is enlisting an influencer army to make it look 'good for the world'","summary":"","url":"https://businessinsider.com/inside-open-ai-influencer-marketing-strategy-chatgpt-ads-sponsorships-instagram-2026-9","source":"Hacker News","publishedAt":"2026-09-23T12:32:00Z","category":"developer","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-376c796fbe4303991789","title":"Show HN: RxFilm Studio–Create and edit your product videos with AI agent","summary":"Hi HN, I built RxFilm Studio, a video workspace that lets you create and revise videos with AI agent. The reason I started this project is simple: I’ve noticed that many independent developers and small teams want to create marketing videos for their product, but producing these videos requires a lot of time and effort","url":"https://filmstudio.rxlab.app/","source":"Hacker News","publishedAt":"2026-09-23T12:28:40Z","category":"developer","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-f9027d80e820682e90d7","title":"Better prompt caching for GPT-6","summary":"Learn how GPT-6 improves prompt caching with higher cache hit rates, new diagnostics, explicit breakpoints, and controls that reduce latency and costs.","url":"https://openai.com/index/better-prompt-caching-for-gpt-6","source":"OpenAI News","publishedAt":"2026-09-22T21:00:00Z","category":"labs","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-a9d177c2bf36d4e504a0","title":"CliffCompaction: Cost-Efficient Compaction for Long-Horizon Coding Agents","summary":"Agents often work on complex problems that require millions of tokens of context, which necessitates compacting across sessions due to limited context windows. We develop CliffCompaction, an autocompaction technique that reduces cost by up to 50% under a bounded context while maintaining or improving performance on Ter","url":"https://arxiv.org/abs/2609.26779v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T17:55:59Z","category":"research","tags":[],"importance":"standard","clusterCount":2},{"id":"sw-1c12f01724c8483d7cc2","title":"The Sirens' Song: When Proximal Background Context Overshadows Distant Evidence","summary":"Long-context LLMs focus on retrieving distant evidence from extensive context, yet existing work has largely focused on overcoming distance alone. In this work, we identify the Proximity Trap, insufficient attention to distant evidence often arises less from distance itself than from cumulative competition with abundan","url":"https://arxiv.org/abs/2609.26718v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T17:06:51Z","category":"research","tags":[],"importance":"standard","clusterCount":2},{"id":"sw-b3a49235b1626f99e56c","title":"The Download: why AI's latest breakthroughs and fears may be more hype than rea","summary":"","url":"https://technologyreview.com/2026/09/22/1144910/the-download-dont-believe-ai-hype","source":"Hacker News","publishedAt":"2026-09-23T11:01:50Z","category":"developer","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-817ee0a5a7a26b8593ed","title":"Train Where the Quantized Model Goes: On-Policy Distillation for Low-Bit Reasoning","summary":"Quantization-aware distillation (QAD) restores much of the short-form question-answering performance lost to sub-3-bit quantization, yet leaves mathematical and code reasoning substantially impaired. Long generations often degenerate into repetitive loops, exhausting the decoding budget without completing a solution. W","url":"https://arxiv.org/abs/2609.26708v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T17:01:09Z","category":"research","tags":[],"importance":"standard","clusterCount":2},{"id":"sw-0680df213917cf3f338b","title":"‘We’re already fighting yesterday’s battle’: Greece’s prime minister gets candid about AI","summary":"Most leaders on a trade mission stick to the pitch, but when I interviewed Greek Prime Minister Kyriakos Mitsotakis this week, he also admitted that no government is ready for what AI is about to do.","url":"https://techcrunch.com/2026/09/22/were-already-fighting-yesterdays-battle-greeces-prime-minister-gets-candid-about-ai","source":"TechCrunch AI","publishedAt":"2026-09-23T04:59:08Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-477260fc0b97bcc11932","title":"Beyond Repeated Sampling: Learning Search Policies for LLM Reasoning","summary":"Large language models increasingly tackle hard reasoning problems by spending more test-time compute, yet the dominant strategy remains naive repeated sampling: draw many independent solutions and hope one is correct. Because such sampling explores only through local decoding noise, it tends to produce many near duplic","url":"https://arxiv.org/abs/2609.26704v1","source":"arXiv CS.CL","publishedAt":"2026-09-22T16:56:53Z","category":"research","tags":[],"importance":"standard","clusterCount":2},{"id":"sw-861ba714619e509b7e39","title":"Measuring the Serving Stack Instead of the Model: Hidden Confounds in Local Tool-Use Evaluation","summary":"A coding agent must emit a valid tool call--a parseable invocation of a tool in the provided schema--before the harness can execute its chosen action. We study how local serving stacks affect this protocol step and show that measured outcomes can depend on the serving layer rather than model behavior alone. In Ollama, ","url":"https://arxiv.org/abs/2609.26693v1","source":"arXiv CS.CL","publishedAt":"2026-09-22T16:48:17Z","category":"research","tags":[],"importance":"standard","clusterCount":2},{"id":"sw-9ae87fb3f5f9f08c326d","title":"A Spectral Theory of Grokking: Weight Decay induces Feature Learning","summary":"In grokking an early fit to the training data separates from a much later improvement in generalization. During this delay, training can move from a fixed neural tangent kernel (NTK) regime to one in which task-relevant kernel eigendirections continue to evolve. We provide a quantitative theory for how this transition ","url":"https://arxiv.org/abs/2609.26679v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T16:38:16Z","category":"research","tags":[],"importance":"standard","clusterCount":2},{"id":"sw-566dfe7d34b559d7ec3b","title":"Faster C++ code intelligence with whole codebase indexing","summary":"C++ code intelligence in GitHub Copilot CLI is now faster with support for whole codebase indexing. C++ repositories can contain millions of lines of code across deeply connected source files… The post Faster C++ code intelligence with whole codebase indexing appeared first on The GitHub Blog.","url":"https://github.blog/changelog/2026-09-22-faster-c-code-intelligence-with-whole-codebase-indexing","source":"GitHub Changelog","publishedAt":"2026-09-22T22:24:48Z","category":"developer","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-abb30a3fd70bf6eb44f0","title":"Discovery-Driven Integration of Disjoint Tables via Text","summary":"Integrating heterogeneous datasets within data lakes is a critical challenge, particularly for semantically related tables that lack the explicit attributes needed to be joined. We study Discovery-Driven Integration, where the relevant sources and their missing relational structure must be discovered before integration","url":"https://arxiv.org/abs/2609.26658v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T16:22:54Z","category":"research","tags":[],"importance":"standard","clusterCount":2},{"id":"sw-9ab67f8ab7e34284341a","title":"The Delegation Blind Spot: Auditing Product Decisions from Agent Choices","summary":"Successful agent execution need not identify which future product improvement its user would value. We present a decision-specific audit that maps a declared observation channel and product-value contrast to compatible intervals and witness populations. Its foundations are established identification and decision theory","url":"https://arxiv.org/abs/2609.26642v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T16:15:55Z","category":"research","tags":[],"importance":"standard","clusterCount":2},{"id":"sw-ed3c4fb0a908e9535100","title":"Capable yet Parsimonious: Extracting and Characterizing Hidden Chain-of-Thought in Frontier Models","summary":"The rapid capability gains of frontier language models are widely attributed to improved reasoning abilities, yet this cannot be verified as raw CoT traces in closed-source systems are hidden. By registering a simple custom tool through a standard API feature, we induce frontier models to externalize intermediate reaso","url":"https://arxiv.org/abs/2609.26637v1","source":"arXiv CS.CL","publishedAt":"2026-09-22T16:10:18Z","category":"research","tags":[],"importance":"standard","clusterCount":2},{"id":"sw-e416bd39fbfa2165fab1","title":"Greedy Decoding Is Not Precision-Invariant: Cross-Precision Output Divergence in LLM Inference","summary":"Greedy decoding from large language models is commonly treated as deterministic. We show it is not precision-invariant: the same model, prompt, and decoding algorithm produce different outputs in BF16 versus FP16 on identical hardware. Across our evaluations of six models (1.1B-7B parameters, four families; divergence ","url":"https://arxiv.org/abs/2609.26621v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T15:54:16Z","category":"research","tags":[],"importance":"standard","clusterCount":2},{"id":"sw-c93d5c29144b12830f3f","title":"Towards Hierarchical GNNs for multi-grid power flow: generalization across operating scenarios","summary":"Hierarchical latent communication improves the generalization of a multi-grid power-flow model to new operating scenarios. The module exchanges information through two reduced graphs within a GENCO-based corrective network. We compare Kron-derived transports, a same-anchor Quotient construction and a flat backbone in p","url":"https://arxiv.org/abs/2609.26603v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T15:41:11Z","category":"research","tags":[],"importance":"standard","clusterCount":2},{"id":"sw-2ec0ad1bc37eb3be3450","title":"Receptiveness, Not Sycophancy: Distinguishing Engagement from Deference in Language Models","summary":"A central concern with language models is sycophancy: their tendency to defer to users' views at the expense of independent substantive judgment. In parallel, work on social sycophancy has focused on behaviors such as validation and positivity that may signal inappropriate deference. Yet the markers of social sycophanc","url":"https://arxiv.org/abs/2609.26579v1","source":"arXiv CS.CL","publishedAt":"2026-09-22T15:27:16Z","category":"research","tags":[],"importance":"standard","clusterCount":2},{"id":"sw-73461b3e97af0d93277a","title":"Introducing GPT-6 Sol and Luna","summary":"Meet GPT-6 Sol and Luna, two models that bring frontier intelligence to everyday work with different balances of capability and cost.","url":"https://openai.com/index/introducing-gpt-6-sol-and-luna","source":"OpenAI News","publishedAt":"2026-09-22T18:00:00Z","category":"labs","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-b340cb36220587db44fb","title":"A Semiotics-Aware Framework for Evaluating Fidelity and Coverage in Natural Language Generation","summary":"When two texts describe the same expression, standard metrics based on lexical overlap or whole-text similarity may fail to detect meaningful differences in how that expression is framed. We propose a framework to evaluate semiotic alignment between texts, where a semiotic profile encompasses both the contextual meanin","url":"https://arxiv.org/abs/2609.26527v1","source":"arXiv CS.CL","publishedAt":"2026-09-22T14:52:06Z","category":"research","tags":[],"importance":"standard","clusterCount":2},{"id":"sw-fdf113d8298f269d4165","title":"Ex-Google safety chief warns AI could harm children more than social media did - Reuters","summary":"Ex-Google safety chief warns AI could harm children more than social media did Reuters","url":"https://news.google.com/rss/articles/CBMiywFBVV95cUxOS3M3dS1XSU9ra3NOc2lNdURNcl9ReGRaRWJXMy01NnJzWHNJNWdYU2FYd2psVjhUSEV2dTNOTjlJZlFJUllzMThDM3RUR1M2VGVBYWw5dVNoaE92SzE3WC12RC1JcWNWNjFkWVJQdGltd01tRl9yb1Z4dzRoMk1zWlRUUUJnczdBQWp0YVBCNThKejZOYUpDUS1YR1gwZlZsZXN5VEhNdFB2Y0JmVjE5RTg4Y1VYSE83RERoM3VxbzI0RE5ZYmtuMndURQ?oc=5","source":"Reuters AI","publishedAt":"2026-09-22T22:45:59Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-24881ef6c8d03e9abaec","title":"Trump says US Justice Department could rein in AI companies if needed - Reuters","summary":"Trump says US Justice Department could rein in AI companies if needed Reuters","url":"https://news.google.com/rss/articles/CBMirgFBVV95cUxOYkFwcnFpOFFycEx0SEwwNUxZazVuTTh6aUNKZ0pUUHlPU2toMkk2WXAyc3RkYTdKWlVsTEVoTmR5VGJaZlhaU21OdHp2aXRWZ0VGNGtJdERKVGUtWXBiMWdKRm9JY04zTnZ3YXp5akdBU2RsS0tHZG9ESTVmR0hOWUYyaWpjYVNCcFF3c3RBVk53NC1SanNTaXpxOHdMei1SNUJPWnhLLTkzU2VOaUE?oc=5","source":"Reuters AI","publishedAt":"2026-09-22T22:32:04Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-07adef8ba424e1ebd0a0","title":"When Recursive Models Finish Computing","summary":"Recursive models can continue updating their latent states beyond their nominal inference budget, so an incorrect output at that budget does not show whether computation is unfinished or has entered a persistently unsuccessful regime. We study the dynamics of completion in attention- and MLP-based Tiny Recursive Models","url":"https://arxiv.org/abs/2609.26487v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T14:24:36Z","category":"research","tags":[],"importance":"standard","clusterCount":2},{"id":"sw-13913172e0a25e8ca804","title":"PP-Net: A Hybrid Physical-Prior Neural Network for Scattered Light Removal in Biomedical Images on Embedded Devices","summary":"Scattered light is common in biomedical images, yet its removal remains challenging. The difficulty arises from three aspects: first, aligned scattered-light-free biomedical ground truth is often unavailable; second, scattering is coupled with weak illumination and sensor-induced noise; and third, many learning-based r","url":"https://arxiv.org/abs/2609.26474v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T14:19:16Z","category":"research","tags":[],"importance":"standard","clusterCount":2},{"id":"sw-cd3f69bbfb2785e56596","title":"Recursive self-improvement of AI research agents","summary":"AI agents are beginning to automate research and development across the AI stack, from improving training efficiency to optimizing inference. A natural next step is to improve the research efficiency of the agents themselves. When an AI research agent's own code is the object of optimization, each accepted rewrite beco","url":"https://arxiv.org/abs/2609.26457v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T14:12:13Z","category":"research","tags":[],"importance":"standard","clusterCount":2},{"id":"sw-0ccb392601be44500a3f","title":"US Treasury secretary could take on AI czar role, source says - Reuters","summary":"US Treasury secretary could take on AI czar role, source says Reuters","url":"https://news.google.com/rss/articles/CBMipAFBVV95cUxQNUpGX2hDdFZBMW45ZFVIelJSUll4TnpWR3NYWExINUxaX2planJ6dU9JNkNCcmM1U216U0VNcDd2TjZfNlluVHUyd1NtZkk2MjZlRTZRRWtLUGJBckpEQk91cUl5QnpvQWc2SC11cTRZczZPa2xBMTBjeVRVMWhRLURoazZZR0E1ZHY4OGNzY0wwTFFCXzRUNFlnbFVmcFFxVjRNRw?oc=5","source":"Reuters AI","publishedAt":"2026-09-22T21:55:00Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-30c310a966d17affe91b","title":"DeepFEAv2: Deep Learning for Transient Finite Element Analysis Beyond Structured Meshes","summary":"Finite Element Analysis (FEA) is widely used for transient mechanical simulations, but its high computational cost limits real-time and high-resolution applications. Deep learning surrogate models can reduce this cost; however, many existing approaches are restricted to steady-state prediction or cannot jointly predict","url":"https://arxiv.org/abs/2609.26426v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T13:51:17Z","category":"research","tags":[],"importance":"standard","clusterCount":2},{"id":"sw-23966244d47d35ba5ef9","title":"v2.1.280","summary":"What's changed Added Claude Opus 5.5 ( claude-opus-5-5 ), now the default Opus model — 1M context, $4/$20 per Mtok with $0.20/Mtok cache reads Added mouse support to more lists in fullscreen mode: the wheel scrolls the /skills list, and a skill's state options in /plugin can be clicked Added CLAUDE_CODE_MAX_MCP_DESCRIP","url":"https://github.com/anthropics/claude-code/releases/tag/v2.1.280","source":"Claude Code Releases","publishedAt":"2026-09-22T16:38:14Z","category":"products","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-9ecc0cb15b595bd1a560","title":"TimeInteract: Towards Real-Time Interactive Intelligence for Streaming Time Series","summary":"Real-world time series evolve continuously, with meaningful changes potentially emerging at any moment. However, existing time-series language models (TSLMs) remain inherently static. They either receive complete sequences for offline processing or alternate between streaming input and response generation, which preven","url":"https://arxiv.org/abs/2609.26389v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T13:29:03Z","category":"research","tags":[],"importance":"standard","clusterCount":2},{"id":"sw-d69ad0fecaa14bf42139","title":"On the Lexical Superstition of Large Language Models for Code Comprehension: Re-evaluation on Code of Low Lexical Quality","summary":"Recent advances in large language models (LLMs) have made them widely used for code-related tasks. Identifier names are statistically informative in naturally occurring code, but their information is not always reliable. We investigate whether current LLMs assign disproportionate weight to lexical cues when renaming pr","url":"https://arxiv.org/abs/2609.26388v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T13:27:57Z","category":"research","tags":[],"importance":"standard","clusterCount":2},{"id":"sw-d11b66e57857c39f0c24","title":"Jev in 25 Lines of Python","summary":"","url":"https://nobodywho.ai/posts/jev-in-25-lines","source":"Hacker News","publishedAt":"2026-09-23T07:26:23Z","category":"developer","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-7ae0263f00a200c4913f","title":"MAVP: Map-Aware Visuomotor Policies for Mobile Manipulation","summary":"Successful mobile manipulation requires coordinated base and arm motion while maintaining accurate spatial positioning. However, demonstration-trained policies can struggle to realise the intended base motion reliably, leading to spatial misalignment and subsequent manipulation failures. We present MAVP (Map-Aware Visu","url":"https://arxiv.org/abs/2609.26378v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T13:22:42Z","category":"research","tags":[],"importance":"standard","clusterCount":2},{"id":"sw-58c39453b24210f6c403","title":"FairMean: Promoting Fairness in Distributed Learning under Label Poisoning Attacks","summary":"Fairness-aware distributed learning prioritizes clients with large losses to reduce performance disparities, but label poisoning can create large losses, thereby inducing a fairness--robustness conflict. We propose FairMean to manage this conflict. FairMean weights client gradients using a bounded, nondecreasing functi","url":"https://arxiv.org/abs/2609.26377v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T13:21:54Z","category":"research","tags":[],"importance":"standard","clusterCount":2},{"id":"sw-27277bfb3ce0fea84c2d","title":"UN chief calls for AI curbs and end to wars in his last assembly address - Reuters","summary":"UN chief calls for AI curbs and end to wars in his last assembly address Reuters","url":"https://news.google.com/rss/articles/CBMiqwFBVV95cUxQTkpSMVRsY0JxTUpDS2t2WDhJNm5pSjdkME96Z2poVEhWM3JnZVU5MDZoMTlFQi1xZ3NZVllGX1JUYWRla0wxdk1aeUFOdEk5VWRZTnhTUmZtZHZ1a1pYNkIwWmpzN1FmaVZOSm5DYWdfTlZQWEZYVEU4OXZXckZ5LVBKVnBWUjhBd0VEQVo0ZkpnT3RkVGc2QUtOcmJlNDdlaEhZZmlYWVRCYWs?oc=5","source":"Reuters AI","publishedAt":"2026-09-22T21:16:50Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-1ae5baf62827f26a7f23","title":"GitScholar: A Dataset for Predicting AI Research Impact from GitHub Engagement","summary":"With the rapid pace of AI research and the hundreds of daily new publications, staying up-to-date with the latest developments has become increasingly difficult. For researchers, quickly identifying impactful work is essential, yet manually reviewing each new publication is impractical. Automated impact prediction meth","url":"https://arxiv.org/abs/2609.26361v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T13:03:31Z","category":"research","tags":[],"importance":"standard","clusterCount":2},{"id":"sw-3cce389477d1dd3c58d0","title":"Flash-dLLM: IO-Aware KV Caching and Parallel Decoding for Fast, Memory-Efficient Diffusion LLMs","summary":"Diffusion Large Language Models (dLLMs) have recently emerged as a promising alternative to autoregressive LLMs by enabling non-autoregressive text generation. However, their practical deployment remains limited by inefficient inference, largely due to the absence of effective Key-Value (KV) caching and scalable parall","url":"https://arxiv.org/abs/2609.26796v1","source":"arXiv CS.CL","publishedAt":"2026-09-22T17:59:57Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-8c799c908d1ca25ebf9b","title":"PACT: From Credit Assignment to Critic Alignment","summary":"Reinforcement learning has become a central component of large language model (LLM) post-training, yet token-level credit lacks a generally accepted mathematical definition, leaving its relationship to commonly used training signals unclear. We formulate three regularity conditions, namely Completeness, Prefix Consiste","url":"https://arxiv.org/abs/2609.26355v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T12:58:28Z","category":"research","tags":[],"importance":"standard","clusterCount":2},{"id":"sw-11815b1966c7e8e3b368","title":"A Decentralized Partially Observable Team Decision Methodology with Delayed Information Sharing","summary":"We study decentralized partially observable team decision problems with low-rank latent dynamics and unknown system models. The proposed framework combines team-theoretic equivalence with low-rank model representations to address cooperative decision-making in partially observable Markov decision processes without prio","url":"https://arxiv.org/abs/2609.26783v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T17:56:46Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-8370f15cad9b252d6b66","title":"Agensh: Scaling Organizational Intelligence to 1,024 Agents","summary":"A multi-agent system can reduce latency on complex tasks by executing work concurrently. Several pioneering harness frameworks support multi-agent systems. However, the scalability of current multi-agent harnesses is often constrained by a central orchestrator's capacity to allocate tasks and coordinate workers. To add","url":"https://arxiv.org/abs/2609.26781v1","source":"arXiv CS.CL","publishedAt":"2026-09-22T17:56:25Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-bad8453607614e3376bd","title":"SWE-Serve: Benchmarking Agentic Engineering For Production Inference Serving","summary":"We introduce SWE-Serve, a benchmark for evaluating agents on production inference engineering tasks. Implementing an inference feature can require coordinating multiple changes across the serving stack, including model support, runtime execution, and public APIs. Existing benchmarks provide limited coverage of producti","url":"https://arxiv.org/abs/2609.26777v1","source":"arXiv CS.AI","publishedAt":"2026-09-22T17:54:59Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-65840b1a4e0d96300081","title":"Geometry-Aware Hyperbolic Residual Quantization","summary":"Residual Vector Quantization turns continuous representations into discrete, multi-level token sequences. Yet most methods operate in Euclidean space, despite the coarse-to-fine structure of the resulting codes and the latent hierarchies present in many data domains. Hyperbolic geometry offers a natural alternative for","url":"https://arxiv.org/abs/2609.26342v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T12:50:18Z","category":"research","tags":[],"importance":"standard","clusterCount":2},{"id":"sw-ce6b87bf0c100a7da2ae","title":"A2M: Trace-Optimized Agent Hijacking in the MCP Ecosystem","summary":"Agents using the Model Context Protocol (MCP) rely on semantic matching to select tools from third-party servers, exposing a semantic supply-chain risk through attacker-controlled metadata and outputs. We introduce A2M (Attraction-to-Manipulation), a two-stage black-box framework for hijacking MCP agents. The Attractio","url":"https://arxiv.org/abs/2609.26761v1","source":"arXiv CS.AI","publishedAt":"2026-09-22T17:41:34Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-aa97e1c7c451cb855aea","title":"Grow the Harness, Not the Context: From Strategy-Free Scaffolds to Reusable Specialist Agents","summary":"Large language model (LLM) agents often handle streams of related tasks, yet standard harnesses repeatedly ask the model to reconstruct the same control decisions inside each task's context. We study whether task feedback can instead turn recurring control into reusable executable code, while reserving LLM calls for ta","url":"https://arxiv.org/abs/2609.26760v1","source":"arXiv CS.AI","publishedAt":"2026-09-22T17:40:45Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-9415b1daa2769fb941b6","title":"Type-Safe Is Not Error-Free: A Constrained Decision Head Follows the Option Name, Not the Rubric Bound to It","summary":"Typed decision models are built for settings where model outputs are consumed directly by software. Instead of generating free-form text, they return a decision over a predefined set of options. By construction, every output conforms to the required schema. Yet this guarantee does not tell us whether the model interpre","url":"https://arxiv.org/abs/2609.26758v1","source":"arXiv CS.AI","publishedAt":"2026-09-22T17:38:12Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-c321be278424670957d0","title":"FleXray: Universal Clinical X-ray Segmentation","summary":"X-ray is medicine's most widely used imaging modality, yet remains among its least quantitative. Unlike volumetric modalities like CT or MRI, X-ray collapses 3D anatomy into a 2D projection, causing structures to overlap and anatomical boundaries to be ambiguous, even to experts. As a result, labeling X-ray databases f","url":"https://arxiv.org/abs/2609.26756v1","source":"arXiv CS.AI","publishedAt":"2026-09-22T17:37:55Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-0a81ae4be51e8f9dc5c3","title":"EquivSVA: A Formally Verified Dataset of Behavioral Assertions Across Equivalent RTL Implementations","summary":"Large language models are increasingly used to generate SystemVerilog Assertions from natural-language specifica- tions and register-transfer-level designs. Existing datasets and benchmarks support important goals such as large- scale training, formal evaluation, specification-to-assertion generation, and mutation-base","url":"https://arxiv.org/abs/2609.26751v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T17:33:09Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-e9fc54b0505e3dd89bfd","title":"Metrics Failure in LLM-Based Code Vulnerability Repair: An Empirical Study and a Change-Aware Screen","summary":"Large language models (LLMs) are increasingly applied to the automated repair of C/C++ security vulnerabilities, and compile rate is a commonly reported proxy for progress: whether the generated patch compiles. We argue that compile rate is a scientifically unreliable metric for single-function vulnerability repair, an","url":"https://arxiv.org/abs/2609.26749v1","source":"arXiv CS.AI","publishedAt":"2026-09-22T17:32:05Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-d715d759e7e7bc56d9da","title":"Automatic depth-based local center clustering via $β$-integrated local depth and adaptive grouping","summary":"Clustering is an unsupervised learning technique that partitions unlabeled data into groups. Most existing methods require user-specified parameters, such as the number of clusters or neighborhood size. Conversely, we propose automatic depth-based local center clustering (A-DLCC), a fully data-driven method that elimin","url":"https://arxiv.org/abs/2609.26748v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T17:31:28Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-9bb3ba5e1fa11514607d","title":"Diffusion-Induced Spatial Attention Overlapping Community Detection","summary":"Detection of overlapping communities is essential for modelling networks in which nodes participate simultaneously in multiple structural or functional groups. Existing graph neural network approaches commonly rely on local message passing, which can obscure community boundaries through smoothing and limit the represen","url":"https://arxiv.org/abs/2609.26737v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T17:25:55Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-ec0e588edf8062bac3b0","title":"Does AI Save Time on Product Design? A Randomized Controlled Experiment of AI Prompt-to-Design Workflows","summary":"AI tools for digital product design now offer prompt-to-design capabilities, allowing designers and their non-designer colleagues to create prototypes through conversational workflows with large language models (LLMs). While these tools promise time savings, experimental evidence in product design remains limited compa","url":"https://arxiv.org/abs/2609.26725v1","source":"arXiv CS.AI","publishedAt":"2026-09-22T17:13:07Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-ef741a249a85dde0e4f4","title":"TraceVIC: Causal Reasoning over Code Evolution for Identifying Vulnerability-Inducing Commits","summary":"Software vulnerabilities are often discovered long after they are introduced, making it difficult to identify the vulnerability-inducing commit (VIC) responsible for introducing the underlying vulnerable condition. Existing VIC identification techniques largely rely on git blame to trace vulnerable code through revisio","url":"https://arxiv.org/abs/2609.26711v1","source":"arXiv CS.AI","publishedAt":"2026-09-22T17:02:42Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-dee9aba0213003f8e4d7","title":"Optimal Sequential Annotations for Off-Policy Evaluation","summary":"Offline reinforcement learning and off-policy evaluation evaluates dynamic treatment rules based on retrospectively collected data prior to deployment. In recent AI applications, state and reward information is recorded as complex text or image, which recent AI advancements such as LLM-as-a-judge can label with unknown","url":"https://arxiv.org/abs/2609.26707v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T17:00:22Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-b61529125db26c651c79","title":"When are bosonic Gaussian states classical to learn?","summary":"A fundamental question in physics is: When does classical behavior emerge from quantum systems? Bosonic Gaussian states provide a natural setting to explore this quantum-classical boundary, as they capture both the classical field behavior and the intrinsic quantum nature of light. Here, we address this problem from a ","url":"https://arxiv.org/abs/2609.26705v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T16:58:25Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-739dcb1469debf2bb3aa","title":"Detecting GPT-Assisted Writing Using Interpretable Stylometric Features","summary":"Distinguishing GPT-assisted from independently authored student writing has become a critical challenge in academia. This paper evaluates the discriminative capability of interpretable stylometric features extracted solely from submitted text. Using data from 90 participants who wrote both independently and with ChatGP","url":"https://arxiv.org/abs/2609.26687v1","source":"arXiv CS.CL","publishedAt":"2026-09-22T16:44:26Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-0d7531998513441bdaa8","title":"PROSWIN: Probabilistic Solar Wind Speed Forecasting Using Deep Distributional Regression From Solar Images","summary":"Accurately predicting fast solar wind conditions is challenging, as uncertainties are large and unquantified by traditional single-value prediction models. In particular, the risks of high-speed solar wind streams (HSSs), which can cause damage to technological infrastructure, cannot be reliably assessed without probab","url":"https://arxiv.org/abs/2609.26683v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T16:42:24Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-8678f1c8ffd73a679401","title":"Woman's brain worm infection confirmed after eggs grow tails in lab test","summary":"There is no single definitive test for these worms, but the tail is fairly convincing.","url":"https://arstechnica.com/health/2026/09/womans-brain-worm-infection-confirmed-after-eggs-grow-tails-in-lab-test","source":"Ars Technica","publishedAt":"2026-09-22T21:42:14Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-91cd13f4c1dac1a6fd5e","title":"From Alignment to Access Control: A Framework for GenAI Policy Enforcement","summary":"Generative AI (GenAI) applications have flourished enabling users to chat with large language models, and to create agents to act on their behalf for a variety of tasks. The pace of development of capabilities in this field is incredibly fast with security and safety taking a back seat. Unfortunately, the slower pace a","url":"https://arxiv.org/abs/2609.26682v1","source":"arXiv CS.AI","publishedAt":"2026-09-22T16:40:57Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-8ddce3d00baaaa5a4867","title":"MAGIC: Mixed-Granularity Agent Graphs via Incremental Construction with Dense-Reward Reinforcement Learning","summary":"Collaboration topology shapes both the performance and execution cost of LLM-based multi-agent systems. Because tasks differ in complexity and required capabilities, recent approaches generate task-specific collaboration graphs that specify agent participation and information flow. However, representative topology gene","url":"https://arxiv.org/abs/2609.26667v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T16:27:46Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-749daf198b9c6f98d866","title":"New Anthropic, OpenAI models make same promise: A little more for a lot less money","summary":"The frontier AI model race has entered its comparison shopping phase.","url":"https://arstechnica.com/ai/2026/09/new-anthropic-openai-models-make-same-promise-a-little-more-for-a-lot-less-money","source":"Ars Technica","publishedAt":"2026-09-22T21:25:18Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-3520c157c4e138eb1c30","title":"TechCrunch Founder Summit’s agenda revealed: Unlock fundraising, hiring, and AI insights in Boston on November 4","summary":"Founders shouldn't have to learn the hardest lessons the hardest way. TechCrunch Founder Summit is designed to make the challenges of starting a company easier and the highs that much greater.","url":"https://techcrunch.com/2026/09/22/techcrunch-founder-summits-agenda-revealed-unlock-fundraising-hiring-and-ai-insights-in-boston-on-november-4","source":"TechCrunch AI","publishedAt":"2026-09-22T23:21:49Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-639acb012e0ed29750d0","title":"OpenAI wants to consult elite mathematicians about how to not fumble again","summary":"After turning a string of spectacular mathematical results into a reputational crisis, OpenAI is consulting human mathematicians to help it figure out a less disastrous path forward. On Monday, the company announced a new independent panel of mathematicians tasked with advising it and other AI companies on their intera","url":"https://theverge.com/ai-artificial-intelligence/999167/openai-elite-mathematicians-panel","source":"The Verge AI","publishedAt":"2026-09-23T00:18:27Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-d7704ea10385c01bd6fb","title":"Statistical Rates for Entropic Optimal Transport in the Discrete to SubGaussian Regime","summary":"We study statistical rates in entropic optimal transport in the semi-discrete regime where one measure has finite support and the other is subGaussian. Our main result establishes parametric convergence rates for the empirical dual potentials to their population counterparts, with no dimension dependence in the leading","url":"https://arxiv.org/abs/2609.26647v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T16:17:58Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-341d5b07e0cf96df5aef","title":"Diffusion Drafts, AR Verifies: Accelerating Document OCR with Self-Speculative Decoding","summary":"Autoregressive OCR vision-language models accurately convert document images into text and structured markup, but require one sequential decoding step per output token, limiting inference speed. Unlike open-ended text generation, OCR outputs are strongly grounded in the input image, making diffusion-based parallel gene","url":"https://arxiv.org/abs/2609.26638v1","source":"arXiv CS.CL","publishedAt":"2026-09-22T16:11:47Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-235d6f2bbf750d6c0d61","title":"Claude Opus 5.5 is now available in GitHub Copilot","summary":"Claude Opus 5.5, Anthropic’s newest Opus model, is now available in GitHub Copilot. You can use it for agentic coding, long-running agentic tasks, and knowledge work. In early testing, Opus… The post Claude Opus 5.5 is now available in GitHub Copilot appeared first on The GitHub Blog.","url":"https://github.blog/changelog/2026-09-22-claude-opus-5-5-is-now-available-in-github-copilot","source":"GitHub Changelog","publishedAt":"2026-09-22T17:10:23Z","category":"developer","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-2f4a5c44304564a49f13","title":"Trump at UN: Threats, diplomacy and 'super intelligence' - Reuters","summary":"Trump at UN: Threats, diplomacy and 'super intelligence' Reuters","url":"https://news.google.com/rss/articles/CBMilAFBVV95cUxPb05jNXUydzFHZ2tRUzBGUndIZFAxTW1jYUtVU3U5NzI1ZjRsckp2YXJuOG9KRE9hOTg1Q3B3b3U1ZC1jWmN1d2J0VU1OMkVfbElFRGhONElJd3I4UklnbVZHTExmT1g2dHFESmJhQWplNWlYUzZsU1ctMW9HRVBXVDJGazNhd1JDbWltQ0hEYmttU2hh?oc=5","source":"Reuters AI","publishedAt":"2026-09-22T19:09:42Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-18449f228374f1495a48","title":"Knowledge Pull Requests for Continual Document Authoring","summary":"We introduce Knowledge Pull Requests (KPRs), a framework for continual document authoring that makes each change interpretable. Documents require ongoing revision as new knowledge surfaces from other sources, languages, or times, but existing approaches either edit with no account of what knowledge changed or regenerat","url":"https://arxiv.org/abs/2609.26634v1","source":"arXiv CS.CL","publishedAt":"2026-09-22T16:07:13Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-403055af7bb31e8e202b","title":"Label-Efficient Learning for Ground-Based Sky-Image Classification: A Benchmark of Transfer Learning, Active Learning, and Pseudo-Labeling on GCD","summary":"Accurate ground-based cloud classification is important for atmospheric monitoring, solar-energy forecasting, aviation weather assessment, and climate observation systems. However, reliable sky-image annotation is time-consuming, especially when cloud types are visually similar or mixed. We study the label efficiency o","url":"https://arxiv.org/abs/2609.26631v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T16:06:22Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-60e478f186f48b1986eb","title":"PERSONAWEAVER: Controllable Diversity Beyond Conventional Archetypes in Procedural Character Generation","summary":"Procedural character generation aims to populate games, simulations, and other virtual worlds with diverse characters. Large language models (LLMs) offer a promising foundation for scaling this task. However, LLM-based procedural character generation remains at an early stage: existing methods either generate character","url":"https://arxiv.org/abs/2609.26629v1","source":"arXiv CS.CL","publishedAt":"2026-09-22T16:04:14Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-f05a574f35fedc0736fa","title":"OpenAI’s GPT-6 Sol and GPT-6 Luna now available","summary":"OpenAI’s GPT-6 family is expanding in GitHub Copilot with two additional models: GPT-6 Sol, and GPT-6 Luna. Joining the previously released GPT-6 Astra, these new options let you select the… The post OpenAI’s GPT-6 Sol and GPT-6 Luna now available appeared first on The GitHub Blog.","url":"https://github.blog/changelog/2026-09-22-openais-gpt-6-sol-and-gpt-6-luna-now-available","source":"GitHub Changelog","publishedAt":"2026-09-22T17:00:14Z","category":"developer","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-6e5daa5d75dfcd53403d","title":"On Basis Function Selection for Sparse Gaussian Process Regression","summary":"Sparse Gaussian processes achieve $O(N)$ inference by replacing the kernel with an appropriate expansion in a fixed basis $\\{φ_j\\}$ on the input space. Given a compute budget $M \\ll N$, practitioners conventionally truncate the basis to its first $M$ entries. Nothing in the formalism, however, prevents one from selecti","url":"https://arxiv.org/abs/2609.26624v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T15:56:54Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-87c823a6b5ca67889ed3","title":"MMAP: Multimodal Missing-Aware Pretraining for Longitudinal Alzheimer's Prediction","summary":"Clinical decision making heavily relies on predicting the disease progression trajectory by seeking to understand patient's health status which is characterised by multimodal medical data. AI holds great potential for learning useful representations from multimodal medical data to predict disease progression and aid cl","url":"https://arxiv.org/abs/2609.26617v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T15:50:42Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-0e0279e89e2d6da97862","title":"Semantic Abstraction for Natural Language Inference: a Methodological Framework for Discovering and Compensating Semantic Knowledge and Reasoning Gaps in Large Language Models","summary":"Despite their outstanding performance on many NLP tasks, LLMs face serious challenges related to semantic abstraction. In this study, we are interested in understanding how LLMs leverage abstract semantic knowledge in natural language inference (NLI), which requires sophisticated linguistic capabilities to interpret im","url":"https://arxiv.org/abs/2609.26610v1","source":"arXiv CS.CL","publishedAt":"2026-09-22T15:46:23Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-91c7b5304d53cac6a1f3","title":"Foundation model embeddings capture pre-diagnostic changes on screening mammograms","summary":"Foundation model embeddings of screening mammograms may encode pre-diagnostic tissue change without task-specific adaptation. We tested whether embeddings move faster along a data-derived \"cancer direction\" in women later biopsied for cancer than in matched screen-negative controls, and whether this depends on pretrain","url":"https://arxiv.org/abs/2609.26605v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T15:42:43Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-9b487a35923c4ef302df","title":"Cities across US oppose Trump FCC plan to preempt local broadband rules","summary":"Cities defend permit requirements, say ISPs aren't building networks fast enough.","url":"https://arstechnica.com/tech-policy/2026/09/cities-across-us-oppose-trump-fcc-plan-to-preempt-local-broadband-rules","source":"Ars Technica","publishedAt":"2026-09-22T20:38:30Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-21df06a31360383700c9","title":"Unlocking Cross-Scenario Physical Layer Security: A Mixture-of-Experts Framework with Generative Diffusion Models","summary":"The future 6G networks are expected to incorporate a proliferation of wireless services in diverse environments, which presents a significant challenge for information security. Conventionally optimization always requires recalculation and learning strategy often suffers poor generalization, which are thus incapable fo","url":"https://arxiv.org/abs/2609.26598v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T15:36:54Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-b9b905c5a7fbb454687a","title":"GTR: Gated Token Recurrence for Efficient Dense Prediction","summary":"Self-attention-based vision backbones perform well on dense prediction, but the quadratic computational cost of global softmax attention limits their efficiency as image resolution increases. We introduce Gated Token Recurrence (GTR), a softmax-free recurrent vision backbone that combines gated linear attention, altern","url":"https://arxiv.org/abs/2609.26590v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T15:34:20Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-7d50e67245d39706d2d6","title":"Polyak-Type Extragradient Methods for Monotone Root-Finding Problems","summary":"We study Polyak-type step-size selection for extragradient methods for solving deterministic and stochastic monotone root-finding problems. We show that the known projection-type correction for deterministic extragradient arises from minimizing an upper bound on the distance to a solution, paralleling the classical Pol","url":"https://arxiv.org/abs/2609.26581v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T15:27:37Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-e8b4866f8702223e3c81","title":"Corporate bond buyers get picky with flood of AI debt - Reuters","summary":"Corporate bond buyers get picky with flood of AI debt Reuters","url":"https://news.google.com/rss/articles/CBMiqgFBVV95cUxNYTBXQ0tEcFNaRTU4dV9GdmJTSmxPVUYzVVhNQmRyWkRHSEx5YTJNMjFLcjNIWFI2R05iSEhSdlF3b29abnEteEcyR3dtWHBCYmNDWmJXcTBqQkdwWUxmWi12Yk1sQWV6VWJPTjZWeG9FckdFT3FlMVhXRkk5SFZyWElxZXVMZ1J2SEpiMm1CZ19jQXVkSnFpUVB6Ujk4cmdyOXNESXhiZXprUQ?oc=5","source":"Reuters AI","publishedAt":"2026-09-22T18:21:10Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-f9d90d43a83470b6f74a","title":"Quantum-Aided Active Device Detection in Energy-Harvesting Symbiotic Radio Networks","summary":"Massive connectivity in next-generation networks demands energy- and spectrum-efficient solutions for large-scale Internet of Things (IoT) deployments. Symbiotic radio (SR) enables passive IoT devices to communicate by backscattering existing cellular transmissions. A key challenge in uplink SR is active device detecti","url":"https://arxiv.org/abs/2609.26565v1","source":"arXiv CS.AI","publishedAt":"2026-09-22T15:17:48Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-1d6e110f9d1d933d2d62","title":"The Disciplinary Language Transfer Problem: How Psychological Vocabulary Produces Governance Failures in AI Agent Deployment","summary":"The vocabulary used to describe AI agents in governance contexts -- learning, memory, values, compliance, identity, trust -- is borrowed from psychological and organizational science, contributing to systematic failures in how organizations deploy, oversee, and hold agents accountable. This paper argues that the proble","url":"https://arxiv.org/abs/2609.26562v1","source":"arXiv CS.AI","publishedAt":"2026-09-22T15:14:36Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-e2503dd9819802ac8144","title":"Neutral-Atom-based Quantum Optimization for Resource Allocation in NOMA Networks","summary":"In wireless communication networks, many resource optimization problems are nondeterministic polynomial-time hard (NP-hard) due to their combinatorial nature and high computational complexity. Recently, neutral-atom-based quantum computing has emerged as a promising platform for efficiently solving such problems by lev","url":"https://arxiv.org/abs/2609.26556v1","source":"arXiv CS.AI","publishedAt":"2026-09-22T15:10:49Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-a9f587cc0c4d552bc40f","title":"JEV-as-a-Judge: Accept When Confident, Escalate When Unsure","summary":"LLM-as-a-judge enables evaluation across diverse tasks, but inference cost and confidence reliability become critical at scale. We study whether a decision-only judge can provide an economical first pass and identify when stronger evaluation is needed. Comparing jev-as-a-judge with sixteen generative and reward-model j","url":"https://arxiv.org/abs/2609.26550v1","source":"arXiv CS.AI","publishedAt":"2026-09-22T15:05:56Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-ed2f9b8f5bc4cd61a115","title":"Topology-Stratified Materials Discovery with A Flow-Based Generative Model","summary":"Accurate generation of crystal structures is the foundation to the discovery of high-performance materials for extreme-environment applications, such as aerospace, additive manufacturing, and fusion energy systems. Although generative modeling has emerged as a promising approach for crystal design, its performance rema","url":"https://arxiv.org/abs/2609.26547v1","source":"arXiv CS.AI","publishedAt":"2026-09-22T15:02:50Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-a67ce36aa13eaede5945","title":"A retrospective analysis on the use of LLMs to study infant syntax learning","summary":"Large language models (LLMs) have increasingly been used to investigate how children acquire syntax at an early stage of development. This is notably the central scientific goal of the BabyLM challenge, a community-wide effort to develop models that achieve human-level syntactic performance while being trained on devel","url":"https://arxiv.org/abs/2609.26539v1","source":"arXiv CS.CL","publishedAt":"2026-09-22T14:58:07Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-f26e12e07358d0c2bafe","title":"Notes on Fourier-Bessel wavelets","summary":"These notes develop the mathematical foundations and construction of a Fourier-Bessel wavelet family inspired by the disk harmonics of Shaqfa et al.[9]. We begin with the relevant properties of Bessel and modified Bessel functions and introduce the wavelet properties required for the construction. We then derive the Fo","url":"https://arxiv.org/abs/2609.26537v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T14:57:54Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-7289dbcceb9e1a5e6288","title":"Transcribe, Translate, and Optimize: Joint Reward Learning for Speech Translation","summary":"In LLM-based speech translation, transcription-based chain-of-thought (CoT) suffers from a mismatch between reference transcripts used in supervised fine-tuning (SFT) and model-generated transcripts at inference. To address this, we propose joint recognition and translation fine-tuning via group relative policy optimiz","url":"https://arxiv.org/abs/2609.26536v1","source":"arXiv CS.CL","publishedAt":"2026-09-22T14:57:51Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-16afc5791c45315aa89d","title":"Snorkel AI triples valuation to $3.5B as demand for AI training data booms","summary":"The seven-year-old startup has raised a $350 million Series E to fuel its data-as-a-service approach.","url":"https://techcrunch.com/2026/09/22/snorkel-ai-triples-valuation-to-3-5b-as-demand-for-ai-training-data-booms","source":"TechCrunch AI","publishedAt":"2026-09-22T21:56:43Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-f801819574dfcc40e8da","title":"REFLEX with Jev for Efficient Selective Control in LLM Agents","summary":"LLM agents often use generative models for bounded decisions, raising the question of when these decisions can be handled more efficiently without reducing task success. We study REFLEX, an agent architecture that uses Jev as a fast, typed decision layer and calls a strong LLM when confidence is low, or generation is r","url":"https://arxiv.org/abs/2609.26532v1","source":"arXiv CS.AI","publishedAt":"2026-09-22T14:54:02Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-b287b34a5390dd7e5b08","title":"Microsoft disrupts AI-assisted platform that compromised 12,000 accounts","summary":"EvilTokens provided an end-to-end platform that makes mass compromises faster and easier.","url":"https://arstechnica.com/security/2026/09/microsoft-disrupts-ai-assisted-platform-that-compromised-12000","source":"Ars Technica","publishedAt":"2026-09-22T19:45:47Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-c8a6aee114e4860049aa","title":"Do Vision Model See Like the Brain? A Comparison Across EEG Encoding Model","summary":"Convolutional neural networks (CNNs) and vision transformers are both used to model the human visual system, but whether the two architectures diverge at a specific point in network depth is unclear. We compared six CNNs and two vision transformers by computing the Pearson correlation (r) between each model's predicted","url":"https://arxiv.org/abs/2609.26512v1","source":"arXiv CS.AI","publishedAt":"2026-09-22T14:41:31Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-35e7e8746d84cd786d85","title":"Gap-Free Streaming PCA Beyond Rank-One Updates: Near-Optimal Rates and Applications to Differential Privacy","summary":"Streaming principal component analysis (PCA) seeks to recover a leading spectral subspace in a single pass over a data stream. We give a new analysis of the ubiquitous Oja's algorithm [Oja82] for the most general, gap-free variant of this problem, where no eigengap assumptions are made on the underlying mean matrix, co","url":"https://arxiv.org/abs/2609.26508v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T14:38:39Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-26d9e15c03aee6ab5267","title":"The Ethics of Artificial Intelligence in Military Operations","summary":"Deep learning systems now mediate military decisions to use force, yet their internal logic resists inspection, their evaluation practices are gameable, and their deployment fractures accountability across dispersed stakeholders. The ethical challenge posed by these systems is fundamentally epistemic: not just whether ","url":"https://arxiv.org/abs/2609.26507v1","source":"arXiv CS.AI","publishedAt":"2026-09-22T14:38:29Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-2ba5cc853eedc5f4fdff","title":"Deep Generative Crystal Structure Prediction: A Benchmark Study and a Controlled Test of Prototype Dependence","summary":"Deep generative models are widely reported to enable de novo crystal structure prediction (CSP), but their capability has not been measured consistently against template-based methods. We evaluate 12 representative generative CSP models, spanning latent-variable, diffusion, flow-matching, autoregressive, and manifold r","url":"https://arxiv.org/abs/2609.26502v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T14:35:38Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-cc002a3849299b187702","title":"Radiomics-Conditioned Modulation of RenalCLIP Features for Clear Cell Renal Cell Carcinoma Classification","summary":"Radiomics provides quantitative descriptions of tumour appearance that may complement disease-specific foundation models in small labelled cohorts. We investigate this complementarity for computed tomography-based classification of clear cell renal cell carcinoma. Our framework uses radiomics to modulate RenalCLIP feat","url":"https://arxiv.org/abs/2609.26492v1","source":"arXiv CS.AI","publishedAt":"2026-09-22T14:29:17Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-b09854e25333af9295fc","title":"Calibration as a First-Class Criterion in LLM Evaluation","summary":"Calibration of language models -- the alignment between expressed or implicit confidence and empirical correctness -- is a well-studied subfield within NLP. Methods to measure it already exist. The problem is adoption: outside this subfield, NLP research regularly introduces new models, datasets, and benchmarks without","url":"https://arxiv.org/abs/2609.26489v1","source":"arXiv CS.CL","publishedAt":"2026-09-22T14:28:31Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-41bbab0be948ec7494a6","title":"Lawsuit demands OpenAI pay for new school after ChatGPT used in shooting","summary":"British Columbia sues OpenAI, demands Tumbler Ridge shooter’s ChatGPT logs.","url":"https://arstechnica.com/tech-policy/2026/09/lawsuit-demands-openai-pay-for-new-school-after-chatgpt-used-in-shooting","source":"Ars Technica","publishedAt":"2026-09-22T19:28:23Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-c5dcad60d7e67fa86b36","title":"Spoken Language Models that Think Aloud","summary":"While Chain-of-Thought (CoT) reasoning has improved the capability of language models, directly applying it to Spoken Language Models (SLMs) may introduce long silent intervals under the serial \"think-then-speak\" paradigm, disrupting real-time spoken interaction. To address this issue, we propose an asynchronous think-","url":"https://arxiv.org/abs/2609.26488v1","source":"arXiv CS.CL","publishedAt":"2026-09-22T14:25:04Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-7d8e8b28203f330fe5de","title":"Not Quite My Tempo: Voice Activity-aware Speech Synthesis for Lip-Synchronous Dubbing","summary":"Automatic lip-synchronous dubbing requires a speech synthesis model to generate alternating voice and silence patterns in the target language that match the timing of the source clip precisely to ensure an optimal viewing experience. Prior works address this problem by conditioning the speech synthesis process on lip m","url":"https://arxiv.org/abs/2609.26486v1","source":"arXiv CS.AI","publishedAt":"2026-09-22T14:24:10Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-7412445894946542eae2","title":"Behavior is Not Enough: A Mechanism-Based Evaluation of Social Norm Emergence in LLM Societies","summary":"Social norms cannot be identified from behavior alone: the same cooperative equilibrium may reflect shared expectations, strategic incentives, or simple imitation. Yet in multi-agent large language model systems, prior work largely treats behavioral convergence as evidence of norm emergence. In this work, we introduce ","url":"https://arxiv.org/abs/2609.26481v1","source":"arXiv CS.CL","publishedAt":"2026-09-22T14:21:04Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-00a6465744fd464b4e8a","title":"FeatLens: Feature-Guided Dynamic Code Graph Construction and Retrieval for Repository-Level Code Generation","summary":"Recent code generation research has moved from isolated function completion toward repository-level generation in existing codebases. To implement a target function correctly, an LLM must identify reusable repository dependencies such as existing functions, APIs, and cross-file definitions. Existing retrieval methods p","url":"https://arxiv.org/abs/2609.26480v1","source":"arXiv CS.AI","publishedAt":"2026-09-22T14:20:46Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-b9ddd6571b9a90773702","title":"How to Estimate Whether You Have Found Several Needles in a Haystack: Measuring Calibration in Multi-Label Text Classification","summary":"A key factor in deciding whether to trust an automatic prediction is its confidence score, which should be calibrated to match the actual probability of the prediction being correct. Most confidence calibration metrics target binary or multi-class tasks, while multi-label calibration remains largely underexplored. Mult","url":"https://arxiv.org/abs/2609.26468v1","source":"arXiv CS.CL","publishedAt":"2026-09-22T14:15:54Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-f54d054fee837e60000d","title":"Complementary Roles of Radiomics and Foundation Representations in Renal Cell Carcinoma Classification: A Comparative Study of 2D and 3D CT Encodings","summary":"Accurate preoperative subtype classification of renal cell carcinoma (RCC) from contrast-enhanced computed tomography remains clinically challenging. Radiomics provides structured tumour descriptors, whereas foundation representations offer transferable image features. However, it remains unclear whether radiomics stil","url":"https://arxiv.org/abs/2609.26463v1","source":"arXiv CS.AI","publishedAt":"2026-09-22T14:14:33Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-a7a7977bb37c74f585a3","title":"Reproducible AI Requires Reproducible Randomness","summary":"Pseudorandom number generators (PRNGs) constitute indispensable computational tools across multiple scientific domains, including Monte Carlo simulations, stochastic computing, and artificial intelligence (AI). The reproducibility of such applications critically depends on the ability of PRNG implementations to generat","url":"https://arxiv.org/abs/2609.26461v1","source":"arXiv CS.AI","publishedAt":"2026-09-22T14:14:18Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-f2d02c87da07f12bc8f6","title":"Can We Predict Anomaly Detection Performance from Embedding-Space Geometry?","summary":"Anomaly detection systems are often trained using normal data alone, while model selection and evaluation typically require labeled anomalies. We study whether anomaly detection performance can be predicted without access to anomalous data. For kNN-based detectors, we derive a lower bound on the area under the ROC curv","url":"https://arxiv.org/abs/2609.26460v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T14:12:54Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-df96110321bb94cae60a","title":"Adobe Premiere finally brings powerful video editing to Android, and it's free","summary":"You don't even need a Creative Cloud login to edit clips, but AI features cost extra.","url":"https://arstechnica.com/gadgets/2026/09/adobe-premiere-finally-brings-powerful-video-editing-to-android-and-its-free","source":"Ars Technica","publishedAt":"2026-09-22T19:08:58Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-b4b5e1002bb549c91ab8","title":"A Practical Guide on Graphical Model Validation","summary":"This manuscript formalizes the most popular model validation tools used in general insurance actuarial modeling. These include graphical tools like calibration plots, actual-vs-expected plots, lift charts, Murphy diagrams, as well as classical statistical tools such as Bregman losses, deviance losses, elementary losses","url":"https://arxiv.org/abs/2609.26445v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T14:04:52Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-2743c18388d485a22f1a","title":"One-Step Generative Surrogate Models via Block-Triangular Joint Drifting","summary":"Drifting provides a direct route to one-step generative models, but applying it directly to stochastic transition modeling requires multiple samples of the next state conditioned on the same current state. Standard trajectory data, however, typically provide only one realized next state for each observed current state ","url":"https://arxiv.org/abs/2609.26435v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T14:00:01Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-15475e1eff6174241981","title":"Siemens Industrial Edge Management","summary":"View CSAF Summary Industrial Edge Management contains an authentication bypass vulnerability that could allow an unauthenticated remote attacker to perform full account takeover by resetting user credentials without completing email verification. Siemens has released new versions for the affected products and recommend","url":"https://cisa.gov/news-events/ics-advisories/icsa-26-265-06","source":"CISA Advisories","publishedAt":"2026-09-22T12:00:00Z","category":"security","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-30618b412412a35c67d9","title":"Siemens Siveillance Control","summary":"View CSAF Summary A vulnerability has been identified in the Open Interface Services (OIS) web module affecting Siveillance Control and Siveillance Control Pro (versions OIS 3.x.y and OIS 4.x.y). This vulnerability allows an attacker to upload arbitrary files, which can lead to unauthorized root-level access on the OIS","url":"https://cisa.gov/news-events/ics-advisories/icsa-26-265-03","source":"CISA Advisories","publishedAt":"2026-09-22T12:00:00Z","category":"security","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-37a83438ae300c55a1b0","title":"Siemens SIPLUS and SIMATIC Products","summary":"View CSAF Summary Multiple Siemens products are vulnerable to the \"Copy Fail\" vulnerability. Siemens has released new versions for several affected products and recommends to update to the latest versions. Siemens is preparing further fix versions and recommends specific countermeasures for products where fixes are not","url":"https://cisa.gov/news-events/ics-advisories/icsa-26-265-04","source":"CISA Advisories","publishedAt":"2026-09-22T12:00:00Z","category":"security","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-3fcefd9c665fb2c24849","title":"Parallel cut research time and cost in half with GPT‑6 Astra","summary":"GPT‑6 Astra allowed Parallel’s agents to research and synthesize labor-market data in half the time and at half the cost vs. prior models.","url":"https://openai.com/index/parallel-cuts-time-and-cost-with-astra","source":"OpenAI News","publishedAt":"2026-09-22T12:00:00Z","category":"labs","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-5b27a0c7066b8fe4465e","title":"Siemens Desigo CC family","summary":"View CSAF Summary A Client Code Execution (CCE) vulnerability has been identified in Desigo CC, potentially allowing malicious actors to execute arbitrary code on client devices through specially crafted graphics documents. This vulnerability leverages user-defined graphics containing embedded scripts that are executed","url":"https://cisa.gov/news-events/ics-advisories/icsa-26-265-05","source":"CISA Advisories","publishedAt":"2026-09-22T12:00:00Z","category":"security","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-5dfa78f335c0ccd58579","title":"Siemens SIMOVE Fleetmanager and SIPLANT","summary":"View CSAF Summary SIMOVE Fleetmanager and SIPLANT contain a path traversal vulnerability that could allow an attacker to access files outside of intended scope. Siemens has released new versions for the affected products and recommends to update to the latest versions. The following versions of Siemens SIMOVE Fleetmana","url":"https://cisa.gov/news-events/ics-advisories/icsa-26-265-07","source":"CISA Advisories","publishedAt":"2026-09-22T12:00:00Z","category":"security","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-707be739d636b0cd6158","title":"Siemens WTV676 and WTV776","summary":"View CSAF Summary The products listed below contain a denial of service vulnerability that could allow an attacker to force the devices into protection mode under certain conditions. This disables remote connectivity functions (Web Access) to the devices. Siemens has released new versions for the affected products and ","url":"https://cisa.gov/news-events/ics-advisories/icsa-26-265-08","source":"CISA Advisories","publishedAt":"2026-09-22T12:00:00Z","category":"security","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-90133b739e3eb2f91daa","title":"lwIP TCP/IP Stack MQTT Client Application","summary":"View CSAF Summary Successful exploitation of this vulnerability could allow an attacker to gain full code execution on the device. The following versions of lwIP TCP/IP Stack MQTT Client Application are affected: MQTT Client Application >=2.0.1|<=2.2.1 (CVE-2026-87121) CVSS Vendor Equipment Vulnerabilities v3 9.8 lwIP ","url":"https://cisa.gov/news-events/ics-advisories/icsa-26-265-01","source":"CISA Advisories","publishedAt":"2026-09-22T12:00:00Z","category":"security","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-9430912d724ffa725aa6","title":"OpenPLC Runtime v3","summary":"View CSAF Summary Successful exploitation of this vulnerability could allow an attacker to hijack session cookies and issue state-changing requests as an operator which would allow the attacker to control the programmable logic controller and the physical processes it drives. The following versions of OpenPLC Runtime v","url":"https://cisa.gov/news-events/ics-advisories/icsa-26-265-09","source":"CISA Advisories","publishedAt":"2026-09-22T12:00:00Z","category":"security","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-d60088c80a69d6f13443","title":"lwIP (Lightweight IP)","summary":"View CSAF Summary Successful exploitation of this vulnerability could result in a system crash, a DoS, or memory corruption, which could lead to code execution on the victim system. The following versions of lwIP (Lightweight IP) are affected: API >=2.0.1|<=2.2.1 (CVE-2026-91018) CVSS Vendor Equipment Vulnerabilities v","url":"https://cisa.gov/news-events/ics-advisories/icsa-26-265-02","source":"CISA Advisories","publishedAt":"2026-09-22T12:00:00Z","category":"security","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-343137e50038640cfae4","title":"The Source of Disturbance Matters: External, Internal, and Control-Generated Noise in Adaptive Regulation","summary":"Adaptive regulation can itself perturb the state it is intended to stabilize. In replicated simulations of an adaptive agent, we compare external disturbance, persistent internally generated disturbance, and control-generated disturbance under regulation-first and disturbance-first ordering. Persistent internal disturb","url":"https://arxiv.org/abs/2609.26428v1","source":"arXiv CS.AI","publishedAt":"2026-09-22T13:53:00Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-a371a57e46831f4e119c","title":"QuantWM: Temporally Consistent 2-Bit KV Cache Quantization for World Models and Video Generation","summary":"KV cache memory has become a major deployment bottleneck for video generation and world models, which motivates low-bit quantization study for efficiency. Existing 2-bit KV cache quantization methods can achieve nearly lossless performance on video benchmarks such as VBench, however, we find that they still cause sever","url":"https://arxiv.org/abs/2609.26425v1","source":"arXiv CS.AI","publishedAt":"2026-09-22T13:50:54Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-f1119370b83df7bd1ca0","title":"Enriching Speech Emotion Representations with Conversational Context","summary":"Detecting emotions is necessary for building systems that can accurately and adaptively interact with humans. Speech Emotion Recognition (SER) has become an important research focus to develop intelligent spoken interfaces. However, most studies predict emotions at the utterance level, ignoring the conversational conte","url":"https://arxiv.org/abs/2609.26422v1","source":"arXiv CS.CL","publishedAt":"2026-09-22T13:50:02Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-a8fd186873dd1ec0529c","title":"Reliability Theory for AI Control","summary":"Reliability theory gives a mature language for layered systems, but its formal tools are not yet standard in frontier AI control. We apply them to Google DeepMind's defenses against rogue deployment. The same control stack can have cubic, quadratic, or linear rare-failure suppression depending on its failure domains. B","url":"https://arxiv.org/abs/2609.26419v1","source":"arXiv CS.AI","publishedAt":"2026-09-22T13:49:46Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-c3b80b62821f10f54fe3","title":"SuperPCA: subspace analysis and an efficient algorithm for high-dimensional PCA","summary":"Principal component analysis (PCA) is a fundamental tool to reduce the dimensionality of the data in many applications. PCA finds a few signal directions that contain most of the variability of the data by computing the eigenvectors of the sample covariance matrix. In this work, we focus on the spiked covariance model,","url":"https://arxiv.org/abs/2609.26406v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T13:39:52Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-eee13521a33207a67cb6","title":"OMatG-flash: An All-Atom Flow Map with Reinforce Adjoint Matching for Scalable Materials Discovery","summary":"The discovery of novel inorganic materials drives technological breakthroughs in critical fields such as computing and energy storage. Generative AI has promised to accelerate the materials discovery pipeline, but state-of-the-art flow and diffusion models remain bottlenecked by the cost of proposing candidate material","url":"https://arxiv.org/abs/2609.26402v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T13:36:33Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-bfb4039aa448be6ffbea","title":"Combining Hierarchical Cognitive Process with Process Supervision for Interpretable Scene Safety Understanding","summary":"Scene safety understanding plays a life-or-death role in situational awareness in various critical domains. Traditional methods that rely on learning direct mappings between scenes and safety levels often lack interpretability, limiting their reliability in critical applications. An effective approach to overcoming thi","url":"https://arxiv.org/abs/2609.26399v1","source":"arXiv CS.CL","publishedAt":"2026-09-22T13:34:53Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-6585cd95e0e327b14284","title":"Double Descent and Malign Overfitting in Diffusion Models","summary":"Conventional wisdom in deep learning holds that overparameterization---having more parameters $p$ than training samples $n$---is benign: larger models generalize better and, even without regularization, interpolating models generalize well, the test error following a double-descent curve. One might expect the same beni","url":"https://arxiv.org/abs/2609.26392v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T13:30:48Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-b5b8c65b47893f05b708","title":"Learning to Defer with Guidance on Real World Medical Data","summary":"Medical image interpretation is high-volume and time-consuming, and while AI interpretation can reduce workload, fully autonomous deployment carries potential safety concerns and low specificity may in practice lead to increased clinician workload. Learning to Defer (L2D) addresses this by selectively routing cases bet","url":"https://arxiv.org/abs/2609.26384v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T13:26:20Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-9ac40b7da3489fc98e9e","title":"Layout-Guided Masking for GROBID: Lightweight Structural Gains in Large-Scale Scientific PDF Ingestion","summary":"Transforming scholarly PDFs into machine-readable fulltext remains a bottleneck for large-scale information systems. Recent vision-based parsers improve accuracy, but need GPUs and may introduce noise into the extracted text. GROBID, a modular font-stream parser running on CPU, is the de-facto standard for structuring ","url":"https://arxiv.org/abs/2609.26381v1","source":"arXiv CS.CL","publishedAt":"2026-09-22T13:24:30Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-d69fb26391d1c9d833b4","title":"Andreessen Horowitz is launching an ‘academy’ with no homework and partnerships with Palantir, Google, and Meta","summary":"Venture capital firm Andreessen Horowitz (a16z) is creating an \"academy\" positioned as a pipeline for young people to build or join a Silicon Valley startup. The \"Horowitz Andreessen Academy\" will launch with 10 partners, including Anduril, Anthropic, Coinbase, Google, Meta, Nvidia, OpenAI, Palantir, Replit, and Stripe","url":"https://theverge.com/ai-artificial-intelligence/998813/andreessen-horowitz-ai-academy","source":"The Verge AI","publishedAt":"2026-09-22T21:22:59Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-01037b2410289b4f9f5f","title":"Security improvements for SSH","summary":"We’re removing several SSH algorithms, adding a new algorithm, and requiring larger RSA SSH keys to improve security. The changes are as follows: We’re removing the ability to use RSA… The post Security improvements for SSH appeared first on The GitHub Blog.","url":"https://github.blog/changelog/2026-09-22-security-improvements-for-ssh","source":"GitHub Changelog","publishedAt":"2026-09-22T14:11:47Z","category":"developer","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-793d1e4b4fd9a8953e4c","title":"HySparse2: Hybrid Sparse Attention with Two-Level KV Sharing","summary":"Long-horizon and multi-turn agents typically generate short actions and process long observations from tools and environments. This growing context demands efficient prefill, compact KV-cache storage, and accurate long-context retrieval. To meet these demands, we introduce HySparse2, a hybrid sparse attention architect","url":"https://arxiv.org/abs/2609.26368v1","source":"arXiv CS.CL","publishedAt":"2026-09-22T13:10:32Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-811bcae574a076a77959","title":"Qualcomm launches two new smartphone chips with emphasis on AI","summary":"Qualcomm said that its new top chip can run 30B mixture-of-expert model locally.","url":"https://techcrunch.com/2026/09/22/qualcomm-launches-two-new-smartphone-chips-with-emphasis-on-ai","source":"TechCrunch AI","publishedAt":"2026-09-22T20:00:00Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-b1bfb607b16ef6fde307","title":"HYDRA: Proactive Android Malware Drift Adaptation via Hierarchical Graph Contrastive Learning","summary":"Concept drift, driven by the rapid evolution of Android malware, severely degrades the performance of machine learning detectors. Current adaptation strategies are often reactive, responding only after performance has dropped and imposing a significant manual annotation burden, or they are proactive but rely on unstabl","url":"https://arxiv.org/abs/2609.26352v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T12:55:25Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-1a05afae6beae4bcfec0","title":"Rabbit’s new AI agent doesn’t need an R1 to run","summary":"Rabbit, the company behind the underwhelming R1 device, is rolling out a standalone AI agent that you don't need its hardware to use, as reported earlier by Wired. The startup says its new OS3 \"agentic operating system\" runs in the cloud but operates locally across Windows, Mac, and Linux devices. According to Rabbit, ","url":"https://theverge.com/ai-artificial-intelligence/999094/rabbit-ai-agent-os3","source":"The Verge AI","publishedAt":"2026-09-22T20:52:40Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-66457c91d404805c2761","title":"Blaming Across the Aisle: Political Contrasting and Blame Attribution in the Danish Parliament","summary":"Political discourse is widely perceived to be growing more hostile, yet robust evidence remains scarce. This study examines blame attribution in the Danish Parliament from 1997 to 2026, combining a purpose-built classifier, BlameBERT (F1: 0.80), with multilevel statistical modeling. The classifier is constructed using ","url":"https://arxiv.org/abs/2609.26346v1","source":"arXiv CS.CL","publishedAt":"2026-09-22T12:52:20Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-c99f9974c1f6be15cd61","title":"Designing and Analysing Argument Mining Pipelines: Towards a Comprehensive Assessment","summary":"Argument Mining (AM) transforms natural language into its underlying argument structures. This transformation is typically realized through a sequence of AM tasks that form an end-to-end AM pipeline. However, AM approaches often differ in how they conceptualize these tasks, making direct comparisons between them diffic","url":"https://arxiv.org/abs/2609.26338v1","source":"arXiv CS.CL","publishedAt":"2026-09-22T12:46:29Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-cfa99734ece5b14f4507","title":"Disaggregated Quantization: Specializing LLM Prefill and Decode","summary":"Prefill and decode reward different approaches to quantization: low-precision arithmetic accelerates prompt processing, while compact weights reduce memory traffic during generation. We propose \"disaggregated quantization\" (DQ), which specializes computation formats, weights and storage placement to both of these phase","url":"https://arxiv.org/abs/2609.26333v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T12:44:16Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-0596b9a3a3f248797717","title":"Error Bounds for Statistical Estimators in BTL Model with Parametric Multivariate Utility Functions","summary":"We study preference elicitation under the Bradley-Terry-Luce (BTL) model where the true partworth vector is unknown and has to be estimated as a parameter with elicited preference information. The set of selected pairwise queries is non-uniform, deterministic, and arbitrary over a collection of alternatives, provided t","url":"https://arxiv.org/abs/2609.26326v1","source":"arXiv CS.LG","publishedAt":"2026-09-22T12:36:59Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-af6c3b10dc422384339c","title":"Review: Resident Evil might just be the best gaming adaptation yet","summary":"Director Zach Cregger's clever reboot strikes perfect balance between horror, humor, and over-the-top gore.","url":"https://arstechnica.com/culture/2026/09/review-resident-evil-is-the-fresh-take-this-flagging-franchise-really-needed","source":"Ars Technica","publishedAt":"2026-09-22T17:30:07Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-c41d2f6d324e74a60cb9","title":"TriWorldBench: A Tri-View Consistency Perspective on Embodied World Models","summary":"Embodied world models predict the outcomes of robot actions to support learning and planning. For robots equipped with head and wrist cameras, this requires complementary views: the head view captures the overall task, while wrist views reveal local gripper-object interactions. However, evaluating these views independe","url":"https://arxiv.org/abs/2609.26314v1","source":"arXiv CS.AI","publishedAt":"2026-09-22T12:23:14Z","category":"research","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-bcba81dc92e5a3e9e40b","title":"Three out of four Americans say AI firms not doing enough to prevent disaster, Reuters/Ipsos poll finds - Reuters","summary":"Three out of four Americans say AI firms not doing enough to prevent disaster, Reuters/Ipsos poll finds Reuters","url":"https://news.google.com/rss/articles/CBMiswFBVV95cUxQX2tkSmZCVGNBY3lkMjNnZHhBT0V0el9sZWdDakE1eHBubVJHdHV3UlVHT1dZT0l0QlVzWXQ5eWEzeVJoUUdhdW8wWS1JTWV2bXBIR0FHaFlpYS1lWjZQRldiOGlibUpOTDI2aTV2X2hRMUw0OEJpcUhZTGUzTUlLbU1kWXNpaTVSUGQxd2JiRE5rNEU0ZEY5WWVEVGtDakp6aEhPR1dsYzZxcW9aWkVuVVczcw?oc=5","source":"Reuters AI","publishedAt":"2026-09-22T15:21:51Z","category":"industry","tags":[],"importance":"standard","clusterCount":1},{"id":"sw-ee284a6ba3961d2bf6e8","title":"Banks warn AI shopping bots raise scam, fraud and data-privacy risks - Reuters","summary":"Banks warn AI shopping bots raise scam, fraud and data-privacy risks Reuters","url":"https://news.google.com/rss/articles/CBMitwFBVV95cUxOUDlYaXB6UFVtWndZUGJsNy1FQzRrM203TUc1RDA0VTZ5Y0FaZ0JHVnV0dzVKMHJkSGowOUFYVUs5N1Y2QWJibzhJZEJpbk8xUmxja1o2ME9tY0hfZjRCOGFGWjVVcmVza2F2VGdvNzJjRnVlY2tzeTdIWWUxREVrNkJYRWRPazhzcG00b0NNYzJDRjlQRmhFV3JlQWNZS2lJWFZfeV9VMmdrdmJiZ0xWNlVkblM2Qk0?oc=5","source":"Reuters AI","publishedAt":"2026-09-22T15:20:52Z","category":"industry","tags":[],"importance":"standard","clusterCount":1}],"sourceHealth":{"healthy":24,"degraded":0,"lastSuccessfulAt":"2026-09-23T20:02:39.062013Z"}}