<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0">
  <channel>
    <title>ainews.now</title>
    <link>https://noisefree.news/</link>
    <description>AI news with the noise removed — act/aware/watch signal, deterministic scrubbing, hourly refresh.</description>
    <language>en</language>
    <lastBuildDate>Tue, 27 Oct 2026 02:14:01 GMT</lastBuildDate>
    <pubDate>Tue, 27 Oct 2026 02:14:01 GMT</pubDate>
    <ttl>60</ttl>
    <item>
      <title>Outbound Investment Security Program Regulations</title>
      <link>https://www.federalregister.gov/documents/2026/10/26/2026-21786/outbound-investment-security-program-regulations</link>
      <guid isPermaLink="true">https://www.federalregister.gov/documents/2026/10/26/2026-21786/outbound-investment-security-program-regulations</guid>
      <pubDate>Mon, 26 Oct 2026 00:00:00 GMT</pubDate>
      <description>This proposed rule sets forth regulations that would implement Subtitle C of the Comprehensive Outbound Investment National Security Act, which is codified at Action: File comments via regulations.gov by Dec 7.</description>
      <category>act</category>
      <category>policy</category>
    </item>
    <item>
      <title>Cooperative Research and Development Agreement: Super Intelligence (SI) Situational Awareness Tools for Analyzing Coast</title>
      <link>https://www.federalregister.gov/documents/2026/10/26/2026-21723/cooperative-research-and-development-agreement-super-intelligence-si-situational-awareness-tools-for</link>
      <guid isPermaLink="true">https://www.federalregister.gov/documents/2026/10/26/2026-21723/cooperative-research-and-development-agreement-super-intelligence-si-situational-awareness-tools-for</guid>
      <pubDate>Mon, 26 Oct 2026 00:00:00 GMT</pubDate>
      <description>The Coast Guard announces its intent to enter a Cooperative Research and Development Agreement (CRADA) with LOOKOUT to assess Super Intelligence for automated Action: File comments via regulations.gov by Nov 25.</description>
      <category>act</category>
      <category>policy</category>
    </item>
    <item>
      <title>llama.cpp b11913</title>
      <link>https://github.com/ggml-org/llama.cpp/releases/tag/b11913</link>
      <guid isPermaLink="true">https://github.com/ggml-org/llama.cpp/releases/tag/b11913</guid>
      <pubDate>Tue, 27 Oct 2026 00:19:01 GMT</pubDate>
      <description>hexagon: add GGML_HEXAGON_VTCM_SIZE (#31221) * hexagon: add GGML_HEXAGON_VTCM_MB and GGML_HEXAGON_NTHREADS caps Change-Id</description>
      <category>aware</category>
      <category>incident</category>
    </item>
    <item>
      <title>ollama v0.40.12-rc0 — v0.40.12</title>
      <link>https://github.com/ollama/ollama/releases/tag/v0.40.12-rc0</link>
      <guid isPermaLink="true">https://github.com/ollama/ollama/releases/tag/v0.40.12-rc0</guid>
      <pubDate>Tue, 27 Oct 2026 00:07:12 GMT</pubDate>
      <description>## What&apos;s Changed * mlx: preserve small-image detail for compact Gemma 4 vision towers by @dhiltgen in https://github.com/ollama/ollama/pull/19247 * MLX</description>
      <category>aware</category>
      <category>open-source</category>
    </item>
    <item>
      <title>GPT-6.1 Terra</title>
      <link>https://simonwillison.net/2026/Oct/26/gpt-61-terra</link>
      <guid isPermaLink="true">https://simonwillison.net/2026/Oct/26/gpt-61-terra</guid>
      <pubDate>Mon, 26 Oct 2026 23:53:07 GMT</pubDate>
      <description>GPT-6.1 Terra It&apos;s always interesting to note which aspects of a new model are emphasized in the release announcement.</description>
      <category>aware</category>
      <category>model-release</category>
    </item>
    <item>
      <title>CVE-2026-109618: RCE via OOM corruption (Update: added to CISA KEV)</title>
      <link>https://cve.now/CVE-2026-109618</link>
      <guid isPermaLink="true">https://cve.now/CVE-2026-109618</guid>
      <pubDate>Mon, 26 Oct 2026 23:09:34 GMT</pubDate>
      <description>Update: added to CISA KEV.</description>
      <category>aware</category>
      <category>incident</category>
    </item>
    <item>
      <title>CVE-2026-109619: RCE via stale global state (Update: added to CISA KEV)</title>
      <link>https://cve.now/CVE-2026-109619</link>
      <guid isPermaLink="true">https://cve.now/CVE-2026-109619</guid>
      <pubDate>Mon, 26 Oct 2026 23:09:34 GMT</pubDate>
      <description>Update: added to CISA KEV.</description>
      <category>aware</category>
      <category>incident</category>
    </item>
    <item>
      <title>langchain langchain-openai==1.7.3</title>
      <link>https://github.com/langchain-ai/langchain/releases/tag/langchain-openai%3D%3D1.7.3</link>
      <guid isPermaLink="true">https://github.com/langchain-ai/langchain/releases/tag/langchain-openai%3D%3D1.7.3</guid>
      <pubDate>Mon, 26 Oct 2026 23:08:48 GMT</pubDate>
      <description>Changes since langchain-openai==1.7.2 test(openai): use current lightweight model in integration tests (#41509) release(openai): 1.7.3 (#41506)</description>
      <category>aware</category>
      <category>incident</category>
    </item>
    <item>
      <title>Cross-request KV cache exposure in the GLM-5.3 Workers AI model (155 responses)</title>
      <link>https://news.ycombinator.com/item?id=50252159</link>
      <guid isPermaLink="true">https://news.ycombinator.com/item?id=50252159</guid>
      <pubDate>Mon, 26 Oct 2026 22:41:48 GMT</pubDate>
      <description>Cross-request KV cache exposure in the GLM-5.3 Workers AI model (155 responses)</description>
      <category>aware</category>
      <category>industry</category>
    </item>
    <item>
      <title>llama.cpp b11912</title>
      <link>https://github.com/ggml-org/llama.cpp/releases/tag/b11912</link>
      <guid isPermaLink="true">https://github.com/ggml-org/llama.cpp/releases/tag/b11912</guid>
      <pubDate>Mon, 26 Oct 2026 22:30:51 GMT</pubDate>
      <description>llama : enable moe amax rescaling for laguna (#31147) * llama : enable moe amax rescaling for laguna Assisted-by: Devin * llama : declare laguna sm tensor</description>
      <category>aware</category>
      <category>open-source</category>
    </item>
    <item>
      <title>Greptile takes bug detection beyond code review with GPT-6 Astra</title>
      <link>https://openai.com/index/greptile-code-review-codex-astra</link>
      <guid isPermaLink="true">https://openai.com/index/greptile-code-review-codex-astra</guid>
      <pubDate>Mon, 26 Oct 2026 22:00:00 GMT</pubDate>
      <description>How Greptile uses GPT-6 Astra to test user workflows, Daybreak to strengthen security, and Codex to reduce rework.</description>
      <category>aware</category>
      <category>industry</category>
    </item>
    <item>
      <title>llama.cpp b11910</title>
      <link>https://github.com/ggml-org/llama.cpp/releases/tag/b11910</link>
      <guid isPermaLink="true">https://github.com/ggml-org/llama.cpp/releases/tag/b11910</guid>
      <pubDate>Mon, 26 Oct 2026 21:20:07 GMT</pubDate>
      <description>opencl: write KV-cache views in set_tensor so a state restore is not dropped (#30471) **Website:** - **Attestations:** - **macOS/iOS:** - [macOS Apple Silicon</description>
      <category>aware</category>
      <category>open-source</category>
    </item>
    <item>
      <title>Proton to &apos;Re-Interview&apos; Staff About AI to See Who Gets to Keep Their Job</title>
      <link>https://news.ycombinator.com/item?id=50251062</link>
      <guid isPermaLink="true">https://news.ycombinator.com/item?id=50251062</guid>
      <pubDate>Mon, 26 Oct 2026 21:13:59 GMT</pubDate>
      <description>Proton to &apos;Re-Interview&apos; Staff About AI to See Who Gets to Keep Their Job</description>
      <category>aware</category>
      <category>industry</category>
    </item>
    <item>
      <title>AI models cost more when you don&apos;t use English</title>
      <link>https://news.ycombinator.com/item?id=50250910</link>
      <guid isPermaLink="true">https://news.ycombinator.com/item?id=50250910</guid>
      <pubDate>Mon, 26 Oct 2026 21:03:22 GMT</pubDate>
      <description>AI models cost more when you don&apos;t use English</description>
      <category>aware</category>
      <category>industry</category>
    </item>
    <item>
      <title>Grand theft AI: Highway thieves are stealing millions in data center gear</title>
      <link>https://news.ycombinator.com/item?id=50250450</link>
      <guid isPermaLink="true">https://news.ycombinator.com/item?id=50250450</guid>
      <pubDate>Mon, 26 Oct 2026 20:28:32 GMT</pubDate>
      <description>Grand theft AI: Highway thieves are stealing millions in data center gear</description>
      <category>aware</category>
      <category>industry</category>
    </item>
    <item>
      <title>Why I&apos;m growing my Vancouver AI company in Taiwan and Europe instead of the U.S.</title>
      <link>https://news.ycombinator.com/item?id=50249953</link>
      <guid isPermaLink="true">https://news.ycombinator.com/item?id=50249953</guid>
      <pubDate>Mon, 26 Oct 2026 19:52:44 GMT</pubDate>
      <description>Why I&apos;m growing my Vancouver AI company in Taiwan and Europe instead of the U.S.</description>
      <category>aware</category>
      <category>industry</category>
    </item>
    <item>
      <title>openai-python v3.39.0</title>
      <link>https://github.com/openai/openai-python/releases/tag/v3.39.0</link>
      <guid isPermaLink="true">https://github.com/openai/openai-python/releases/tag/v3.39.0</guid>
      <pubDate>Mon, 26 Oct 2026 19:14:00 GMT</pubDate>
      <description>## [3.39.0](https://github.com/openai/openai-python/compare/v3.38.0...v3.39.0) (2026-10-26) ### Features * **api:** add gpt-6.1-terra model identifier</description>
      <category>aware</category>
      <category>open-source</category>
    </item>
    <item>
      <title>Quoting Jason Koebler</title>
      <link>https://simonwillison.net/2026/Oct/26/jason-koebler</link>
      <guid isPermaLink="true">https://simonwillison.net/2026/Oct/26/jason-koebler</guid>
      <pubDate>Mon, 26 Oct 2026 19:00:25 GMT</pubDate>
      <description>Because big tech platforms essentially do not have customer support and make it impossible for their users to contact a human, many people experiencing</description>
      <category>aware</category>
      <category>incident</category>
    </item>
    <item>
      <title>GPT-6.1 Terra (Max) Scores 48 on Artificial Analysis</title>
      <link>https://news.ycombinator.com/item?id=50249076</link>
      <guid isPermaLink="true">https://news.ycombinator.com/item?id=50249076</guid>
      <pubDate>Mon, 26 Oct 2026 18:54:00 GMT</pubDate>
      <description>GPT-6.1 Terra (Max) Scores 48 on Artificial Analysis</description>
      <category>aware</category>
      <category>research</category>
    </item>
    <item>
      <title>transformers v5.21</title>
      <link>https://github.com/huggingface/transformers/releases/tag/v5.21</link>
      <guid isPermaLink="true">https://github.com/huggingface/transformers/releases/tag/v5.21</guid>
      <pubDate>Mon, 26 Oct 2026 18:17:42 GMT</pubDate>
      <description># Release v5.21.0 ## New Model additions ### MiniCPM-OCR MiniCPM-OCR is a document-understanding model from OpenBMB&apos;s MiniCPM-V family.</description>
      <category>aware</category>
      <category>open-source</category>
    </item>
    <item>
      <title>Concerning behavior from a popular open-weight AI model</title>
      <link>https://news.ycombinator.com/item?id=50246658</link>
      <guid isPermaLink="true">https://news.ycombinator.com/item?id=50246658</guid>
      <pubDate>Mon, 26 Oct 2026 16:35:44 GMT</pubDate>
      <description>Concerning behavior from a popular open-weight AI model</description>
      <category>aware</category>
      <category>industry</category>
    </item>
    <item>
      <title>openai-python v3.38.0</title>
      <link>https://github.com/openai/openai-python/releases/tag/v3.38.0</link>
      <guid isPermaLink="true">https://github.com/openai/openai-python/releases/tag/v3.38.0</guid>
      <pubDate>Mon, 26 Oct 2026 16:35:03 GMT</pubDate>
      <description>## [3.38.0](https://github.com/openai/openai-python/compare/v3.37.0...v3.38.0) (2026-10-25) ### Features * **api:** add delegation controls and prompt cache</description>
      <category>aware</category>
      <category>incident</category>
    </item>
    <item>
      <title>webframe: evaluated as a scraper and a sandbox</title>
      <link>https://simonwillison.net/2026/Oct/26/webframe-evaluation</link>
      <guid isPermaLink="true">https://simonwillison.net/2026/Oct/26/webframe-evaluation</guid>
      <pubDate>Mon, 26 Oct 2026 16:15:00 GMT</pubDate>
      <description>Research: webframe: evaluated as a scraper and a sandbox webframe is a neat little project from Jeff Lindsay which lets you &quot;run a web page like a program&quot; -</description>
      <category>aware</category>
      <category>open-source</category>
    </item>
    <item>
      <title>Show HN: Hob - Open-source, on device alternative to Claude Cowork and ChatGPT</title>
      <link>https://news.ycombinator.com/item?id=50246238</link>
      <guid isPermaLink="true">https://news.ycombinator.com/item?id=50246238</guid>
      <pubDate>Mon, 26 Oct 2026 16:12:29 GMT</pubDate>
      <description>Show HN: Hob - Open-source, on device alternative to Claude Cowork and ChatGPT</description>
      <category>aware</category>
      <category>open-source</category>
    </item>
    <item>
      <title>Working with academic communities to build scientific understanding</title>
      <link>https://openai.com/index/building-scientific-understanding-with-academic-communities</link>
      <guid isPermaLink="true">https://openai.com/index/building-scientific-understanding-with-academic-communities</guid>
      <pubDate>Mon, 26 Oct 2026 14:30:00 GMT</pubDate>
      <description>OpenAI shares model results on the Kakeya conjecture, Exponential Time Hypothesis, and quantum chromodynamics—and how researchers can verify and build on them.</description>
      <category>aware</category>
      <category>industry</category>
    </item>
    <item>
      <title>CVE-2026-107656: An authentication bypass using an alternate path or channel vulnerability in Fortinet FortiSandbox</title>
      <link>https://cve.now/CVE-2026-107656</link>
      <guid isPermaLink="true">https://cve.now/CVE-2026-107656</guid>
      <pubDate>Mon, 26 Oct 2026 14:01:34 GMT</pubDate>
      <description>Fortinet FortiSandbox and FortiSandbox PaaS are affected by an authentication bypass flaw that lets an attacker reach protected functions through an alternate</description>
      <category>aware</category>
      <category>incident</category>
    </item>
    <item>
      <title>GPT-6.1 Sol helps CodeRabbit strengthen security across complex codebases</title>
      <link>https://openai.com/index/gpt-6-1-sol-coderabbit-code-security</link>
      <guid isPermaLink="true">https://openai.com/index/gpt-6-1-sol-coderabbit-code-security</guid>
      <pubDate>Mon, 26 Oct 2026 13:45:00 GMT</pubDate>
      <description>CodeRabbit uses GPT-6.1 Sol to investigate security risks across complex codebases.</description>
      <category>aware</category>
      <category>industry</category>
    </item>
    <item>
      <title>Brown student maps the hidden global labor powering AI</title>
      <link>https://news.ycombinator.com/item?id=50242843</link>
      <guid isPermaLink="true">https://news.ycombinator.com/item?id=50242843</guid>
      <pubDate>Mon, 26 Oct 2026 12:58:27 GMT</pubDate>
      <description>Brown student maps the hidden global labor powering AI</description>
      <category>aware</category>
      <category>industry</category>
    </item>
    <item>
      <title>Import AI 478: Math for alignment; ProgramBench; and Mirendil shows what RSI means</title>
      <link>https://importai.substack.com/p/import-ai-478-math-for-alignment</link>
      <guid isPermaLink="true">https://importai.substack.com/p/import-ai-478-math-for-alignment</guid>
      <pubDate>Mon, 26 Oct 2026 12:31:52 GMT</pubDate>
      <description>Import AI 478: Math for alignment; ProgramBench; and Mirendil shows what RSI means</description>
      <category>aware</category>
      <category>industry</category>
    </item>
    <item>
      <title>ChatGPT’s expanding world of work</title>
      <link>https://openai.com/index/chatgpt-expanding-world-of-work</link>
      <guid isPermaLink="true">https://openai.com/index/chatgpt-expanding-world-of-work</guid>
      <pubDate>Mon, 26 Oct 2026 10:00:00 GMT</pubDate>
      <description>Explore OpenAI Signals data on 165 work activities across nearly 100 countries, revealing how people use ChatGPT for work and how usage changes over time.</description>
      <category>aware</category>
      <category>industry</category>
    </item>
    <item>
      <title>How Fanatics and OpenAI streamline the fan experience</title>
      <link>https://openai.com/index/fanatics-streamline-fan-experience</link>
      <guid isPermaLink="true">https://openai.com/index/fanatics-streamline-fan-experience</guid>
      <pubDate>Mon, 26 Oct 2026 07:00:00 GMT</pubDate>
      <description>Fanatics uses OpenAI to help fans complete their game-day outfits with Shop the Look.</description>
      <category>aware</category>
      <category>industry</category>
    </item>
    <item>
      <title>How Ultrafast cut incident investigation from hours to minutes at OpenAI</title>
      <link>https://openai.com/index/openai-cut-incident-investigation</link>
      <guid isPermaLink="true">https://openai.com/index/openai-cut-incident-investigation</guid>
      <pubDate>Mon, 26 Oct 2026 07:00:00 GMT</pubDate>
      <description>See how OpenAI engineers use Codex Ultrafast to investigate incidents in minutes.</description>
      <category>aware</category>
      <category>incident</category>
    </item>
    <item>
      <title>Cognitive Mercantilism: AI panic resembles fears of early capitalism</title>
      <link>https://news.ycombinator.com/item?id=50239027</link>
      <guid isPermaLink="true">https://news.ycombinator.com/item?id=50239027</guid>
      <pubDate>Mon, 26 Oct 2026 04:11:40 GMT</pubDate>
      <description>Cognitive Mercantilism: AI panic resembles fears of early capitalism</description>
      <category>aware</category>
      <category>industry</category>
    </item>
    <item>
      <title>If AI displaces human workers, they will lose more than just pay</title>
      <link>https://news.ycombinator.com/item?id=50238351</link>
      <guid isPermaLink="true">https://news.ycombinator.com/item?id=50238351</guid>
      <pubDate>Mon, 26 Oct 2026 02:22:38 GMT</pubDate>
      <description>If AI displaces human workers, they will lose more than just pay</description>
      <category>aware</category>
      <category>industry</category>
    </item>
    <item>
      <title>Introducing GPT-6.1 Terra</title>
      <link>https://openai.com/index/introducing-gpt-6-1-terra</link>
      <guid isPermaLink="true">https://openai.com/index/introducing-gpt-6-1-terra</guid>
      <pubDate>Mon, 26 Oct 2026 00:00:00 GMT</pubDate>
      <description>Introducing GPT-6.1 Terra</description>
      <category>aware</category>
      <category>model-release</category>
    </item>
    <item>
      <title>Apollo says AI has delivered &apos;zero financial payoff&apos; outside the tech industry</title>
      <link>https://news.ycombinator.com/item?id=50236828</link>
      <guid isPermaLink="true">https://news.ycombinator.com/item?id=50236828</guid>
      <pubDate>Sun, 25 Oct 2026 22:49:06 GMT</pubDate>
      <description>Apollo says AI has delivered &apos;zero financial payoff&apos; outside the tech industry</description>
      <category>aware</category>
      <category>industry</category>
    </item>
    <item>
      <title>OpenAI and Anthropic cut their LLM prices by up to 80% within a few months. Why?</title>
      <link>https://news.ycombinator.com/item?id=50236102</link>
      <guid isPermaLink="true">https://news.ycombinator.com/item?id=50236102</guid>
      <pubDate>Sun, 25 Oct 2026 21:11:31 GMT</pubDate>
      <description>OpenAI and Anthropic cut their LLM prices by up to 80% within a few months. Why?</description>
      <category>aware</category>
      <category>industry</category>
    </item>
    <item>
      <title>AI didn&apos;t cause the tech jobs bust, but it explains why there&apos;s no recovery</title>
      <link>https://news.ycombinator.com/item?id=50234699</link>
      <guid isPermaLink="true">https://news.ycombinator.com/item?id=50234699</guid>
      <pubDate>Sun, 25 Oct 2026 18:42:23 GMT</pubDate>
      <description>AI didn&apos;t cause the tech jobs bust, but it explains why there&apos;s no recovery</description>
      <category>aware</category>
      <category>industry</category>
    </item>
    <item>
      <title>CVE-2026-111497: Out-of-Bounds Write in Google Brotli Decoder (Update: now rated critical)</title>
      <link>https://cve.now/CVE-2026-111497</link>
      <guid isPermaLink="true">https://cve.now/CVE-2026-111497</guid>
      <pubDate>Sun, 25 Oct 2026 17:37:34 GMT</pubDate>
      <description>Update: now rated critical.</description>
      <category>aware</category>
      <category>incident</category>
    </item>
    <item>
      <title>CVE-2026-45484: Microsoft SharePoint Elevation of Privilege Vulnerability (Update: now rated critical)</title>
      <link>https://cve.now/CVE-2026-45484</link>
      <guid isPermaLink="true">https://cve.now/CVE-2026-45484</guid>
      <pubDate>Sun, 25 Oct 2026 17:37:34 GMT</pubDate>
      <description>Update: now rated critical.</description>
      <category>aware</category>
      <category>incident</category>
    </item>
    <item>
      <title>CVE-2026-45657: Windows Kernel Remote Code Execution Vulnerability (Update: now rated critical)</title>
      <link>https://cve.now/CVE-2026-45657</link>
      <guid isPermaLink="true">https://cve.now/CVE-2026-45657</guid>
      <pubDate>Sun, 25 Oct 2026 17:37:34 GMT</pubDate>
      <description>Update: now rated critical.</description>
      <category>aware</category>
      <category>incident</category>
    </item>
    <item>
      <title>CVE-2026-47291: HTTP.sys Remote Code Execution Vulnerability (Update: now rated critical)</title>
      <link>https://cve.now/CVE-2026-47291</link>
      <guid isPermaLink="true">https://cve.now/CVE-2026-47291</guid>
      <pubDate>Sun, 25 Oct 2026 17:37:34 GMT</pubDate>
      <description>Update: now rated critical.</description>
      <category>aware</category>
      <category>incident</category>
    </item>
    <item>
      <title>CVE-2026-49160: HTTP.sys Denial of Service Vulnerability (Update: now rated critical)</title>
      <link>https://cve.now/CVE-2026-49160</link>
      <guid isPermaLink="true">https://cve.now/CVE-2026-49160</guid>
      <pubDate>Sun, 25 Oct 2026 17:37:34 GMT</pubDate>
      <description>Update: now rated critical.</description>
      <category>aware</category>
      <category>incident</category>
    </item>
    <item>
      <title>CVE-2026-50656: Microsoft Defender Elevation of Privilege Vulnerability (Update: now rated critical)</title>
      <link>https://cve.now/CVE-2026-50656</link>
      <guid isPermaLink="true">https://cve.now/CVE-2026-50656</guid>
      <pubDate>Sun, 25 Oct 2026 17:37:34 GMT</pubDate>
      <description>Update: now rated critical.</description>
      <category>aware</category>
      <category>incident</category>
    </item>
    <item>
      <title>CVE-2026-40372: ASP.NET Core Elevation of Privilege Vulnerability (Update: now rated critical)</title>
      <link>https://cve.now/CVE-2026-40372</link>
      <guid isPermaLink="true">https://cve.now/CVE-2026-40372</guid>
      <pubDate>Sun, 25 Oct 2026 17:37:34 GMT</pubDate>
      <description>Update: now rated critical.</description>
      <category>aware</category>
      <category>incident</category>
    </item>
    <item>
      <title>CVE-2026-17176: OS Command Injection Vulnerability in Deco Devices (Update: now rated critical)</title>
      <link>https://cve.now/CVE-2026-17176</link>
      <guid isPermaLink="true">https://cve.now/CVE-2026-17176</guid>
      <pubDate>Sun, 25 Oct 2026 17:37:34 GMT</pubDate>
      <description>Update: now rated critical.</description>
      <category>aware</category>
      <category>incident</category>
    </item>
    <item>
      <title>CVE-2026-93674: Langflow OSS is affected by multiple vulnerabilities (Update: now rated critical)</title>
      <link>https://cve.now/CVE-2026-93674</link>
      <guid isPermaLink="true">https://cve.now/CVE-2026-93674</guid>
      <pubDate>Sun, 25 Oct 2026 17:37:34 GMT</pubDate>
      <description>Update: now rated critical.</description>
      <category>aware</category>
      <category>incident</category>
    </item>
    <item>
      <title>CVE-2026-46300: net: skbuff: preserve shared-frag marker during coalescing (Update: now rated critical)</title>
      <link>https://cve.now/CVE-2026-46300</link>
      <guid isPermaLink="true">https://cve.now/CVE-2026-46300</guid>
      <pubDate>Sun, 25 Oct 2026 17:37:34 GMT</pubDate>
      <description>Update: now rated critical.</description>
      <category>aware</category>
      <category>incident</category>
    </item>
    <item>
      <title>CVE-2026-4631: Cockpit: cockpit: unauthenticated remote code execution due to ssh command-line argument injection</title>
      <link>https://cve.now/CVE-2026-4631</link>
      <guid isPermaLink="true">https://cve.now/CVE-2026-4631</guid>
      <pubDate>Sun, 25 Oct 2026 17:37:34 GMT</pubDate>
      <description>Update: now rated critical.</description>
      <category>aware</category>
      <category>incident</category>
    </item>
    <item>
      <title>A 10-year-old CNN outperforms frontier LLMs for skin cancer diagnosis</title>
      <link>https://news.ycombinator.com/item?id=50233779</link>
      <guid isPermaLink="true">https://news.ycombinator.com/item?id=50233779</guid>
      <pubDate>Sun, 25 Oct 2026 17:15:38 GMT</pubDate>
      <description>A 10-year-old CNN outperforms frontier LLMs for skin cancer diagnosis</description>
      <category>aware</category>
      <category>research</category>
    </item>
    <item>
      <title>AI use leads UC Berkeley math professor to see &apos;no point&apos; in grading homework</title>
      <link>https://news.ycombinator.com/item?id=50233304</link>
      <guid isPermaLink="true">https://news.ycombinator.com/item?id=50233304</guid>
      <pubDate>Sun, 25 Oct 2026 16:28:26 GMT</pubDate>
      <description>AI use leads UC Berkeley math professor to see &apos;no point&apos; in grading homework</description>
      <category>aware</category>
      <category>industry</category>
    </item>
    <item>
      <title>Seam Cutter</title>
      <link>https://simonwillison.net/2026/Oct/25/seam-cutter</link>
      <guid isPermaLink="true">https://simonwillison.net/2026/Oct/25/seam-cutter</guid>
      <pubDate>Sun, 25 Oct 2026 15:10:19 GMT</pubDate>
      <description>Tool: Seam Cutter I built this weird and very specific cropping tool - it lets you create a horizontal or vertical cut across an image, then slide the two</description>
      <category>aware</category>
      <category>open-source</category>
    </item>
    <item>
      <title>How to do method research when I feel I have no guidance on doing it? [D]</title>
      <link>https://www.reddit.com/r/MachineLearning/comments/1xgdl42/how_to_do_method_research_when_i_feel_i_have_no</link>
      <guid isPermaLink="true">https://www.reddit.com/r/MachineLearning/comments/1xgdl42/how_to_do_method_research_when_i_feel_i_have_no</guid>
      <pubDate>Tue, 27 Oct 2026 00:06:43 GMT</pubDate>
      <description>I am a 3rd year PhD student in a small lab working on ML in healthcare.</description>
      <category>watch</category>
      <category>industry</category>
    </item>
    <item>
      <title>How to patch PHP2Shell with free Aikido Images</title>
      <link>https://www.aikido.dev/blog/how-to-patch-php2shell-with-free-aikido-images</link>
      <guid isPermaLink="true">https://www.aikido.dev/blog/how-to-patch-php2shell-with-free-aikido-images</guid>
      <pubDate>Mon, 26 Oct 2026 22:15:00 GMT</pubDate>
      <description>Patch PHP2Shell (CVE-2026-109618 and CVE-2026-109619) with free Aikido Images for PHP 5.5 to 8.5.</description>
      <category>watch</category>
      <category>incident</category>
    </item>
    <item>
      <title>Deepity: a hand-tuned C++/CUDA predictive coding library, beats jpc (JAX) and ngc-learn on MNIST (CPU benchmarks +</title>
      <link>https://www.reddit.com/r/MachineLearning/comments/1xg6sqa/deepity_a_handtuned_ccuda_predictive_coding</link>
      <guid isPermaLink="true">https://www.reddit.com/r/MachineLearning/comments/1xg6sqa/deepity_a_handtuned_ccuda_predictive_coding</guid>
      <pubDate>Mon, 26 Oct 2026 19:34:57 GMT</pubDate>
      <description>I&apos;ve been building a predictive coding (PC) engine from scratch in C++20, with a thin Python layer on top, and just cut the first public release.</description>
      <category>watch</category>
      <category>open-source</category>
    </item>
    <item>
      <title>Fitting a 4D Mandelbulb with KS3 vs MLP [R]</title>
      <link>https://www.reddit.com/r/MachineLearning/comments/1xg36lw/fitting_a_4d_mandelbulb_with_ks3_vs_mlp_r</link>
      <guid isPermaLink="true">https://www.reddit.com/r/MachineLearning/comments/1xg36lw/fitting_a_4d_mandelbulb_with_ks3_vs_mlp_r</guid>
      <pubDate>Mon, 26 Oct 2026 17:22:39 GMT</pubDate>
      <description>For the past year I have been studying using a fast K-dim hermite spline manifold algorithm, called K-Splanifolds or KS, for general purpose fast regression.</description>
      <category>watch</category>
      <category>industry</category>
    </item>
    <item>
      <title>Scoring IDP [R]</title>
      <link>https://www.reddit.com/r/MachineLearning/comments/1xg1695/scoring_idp_r</link>
      <guid isPermaLink="true">https://www.reddit.com/r/MachineLearning/comments/1xg1695/scoring_idp_r</guid>
      <pubDate>Mon, 26 Oct 2026 16:10:06 GMT</pubDate>
      <description>This is a &quot;survey of current practices&quot; type of research for a niche area, nothing earth-shattering.</description>
      <category>watch</category>
      <category>research</category>
    </item>
    <item>
      <title>PBT doesn’t make grokking faster, but it can escape a hyperparameter range where nothing groks [P]</title>
      <link>https://www.reddit.com/r/MachineLearning/comments/1xfqhfc/pbt_doesnt_make_grokking_faster_but_it_can_escape</link>
      <guid isPermaLink="true">https://www.reddit.com/r/MachineLearning/comments/1xfqhfc/pbt_doesnt_make_grokking_faster_but_it_can_escape</guid>
      <pubDate>Mon, 26 Oct 2026 07:48:07 GMT</pubDate>
      <description>I ran Population Based Training over learning rate and weight decay on grokking (modular addition mod 97, 35% train split, small MLP, 20 members, 20k steps</description>
      <category>watch</category>
      <category>research</category>
    </item>
    <item>
      <title>I ran a blind test on self-hosted AI agents. Here are the results, guess who&apos;s who [D]</title>
      <link>https://www.reddit.com/r/MachineLearning/comments/1xfgf1w/i_ran_a_blind_test_on_selfhosted_ai_agents_here</link>
      <guid isPermaLink="true">https://www.reddit.com/r/MachineLearning/comments/1xfgf1w/i_ran_a_blind_test_on_selfhosted_ai_agents_here</guid>
      <pubDate>Sun, 25 Oct 2026 22:57:33 GMT</pubDate>
      <description>everything here ran on my own hardware, an M5 Max MacBook Pro with 64GB.</description>
      <category>watch</category>
      <category>open-source</category>
    </item>
    <item>
      <title>What&apos;s the hardest part of creating edge cases for fine-tuning open-weight models? [D]</title>
      <link>https://www.reddit.com/r/MachineLearning/comments/1xfa17z/whats_the_hardest_part_of_creating_edge_cases_for</link>
      <guid isPermaLink="true">https://www.reddit.com/r/MachineLearning/comments/1xfa17z/whats_the_hardest_part_of_creating_edge_cases_for</guid>
      <pubDate>Sun, 25 Oct 2026 18:35:21 GMT</pubDate>
      <description>I&apos;m interested in how people build datasets for fine-tuning open-weight models, especially when it is about making a model handling edge cases than just</description>
      <category>watch</category>
      <category>open-source</category>
    </item>
    <item>
      <title>Nanbeige releases Nanbeige4.3-33B, an Apache 2.0 teacher model for smaller agents [N]</title>
      <link>https://www.reddit.com/r/MachineLearning/comments/1xf6x35/nanbeige_releases_nanbeige4333b_an_apache_20</link>
      <guid isPermaLink="true">https://www.reddit.com/r/MachineLearning/comments/1xf6x35/nanbeige_releases_nanbeige4333b_an_apache_20</guid>
      <pubDate>Sun, 25 Oct 2026 16:32:44 GMT</pubDate>
      <description>Nanbeige released Nanbeige4.3-33B as the teacher for its planned smaller agents, including an upcoming 3B model.</description>
      <category>watch</category>
      <category>model-release</category>
    </item>
    <item>
      <title>A mathematician reads the AI future</title>
      <link>https://news.ycombinator.com/item?id=50233347</link>
      <guid isPermaLink="true">https://news.ycombinator.com/item?id=50233347</guid>
      <pubDate>Sun, 25 Oct 2026 16:31:21 GMT</pubDate>
      <description>A mathematician reads the AI future</description>
      <category>watch</category>
      <category>industry</category>
    </item>
    <item>
      <title>R/open source bans all solo projects and projects using LLMs</title>
      <link>https://news.ycombinator.com/item?id=50231722</link>
      <guid isPermaLink="true">https://news.ycombinator.com/item?id=50231722</guid>
      <pubDate>Sun, 25 Oct 2026 14:05:03 GMT</pubDate>
      <description>R/open source bans all solo projects and projects using LLMs</description>
      <category>watch</category>
      <category>open-source</category>
    </item>
    <item>
      <title>Imagine an AI future where being banned means you are now economically useless</title>
      <link>https://news.ycombinator.com/item?id=50230995</link>
      <guid isPermaLink="true">https://news.ycombinator.com/item?id=50230995</guid>
      <pubDate>Sun, 25 Oct 2026 12:47:20 GMT</pubDate>
      <description>Imagine an AI future where being banned means you are now economically useless</description>
      <category>watch</category>
      <category>industry</category>
    </item>
    <item>
      <title>Nikon photography contest winner loses title after scientists spot signs of AI</title>
      <link>https://news.ycombinator.com/item?id=50229827</link>
      <guid isPermaLink="true">https://news.ycombinator.com/item?id=50229827</guid>
      <pubDate>Sun, 25 Oct 2026 09:51:17 GMT</pubDate>
      <description>Nikon photography contest winner loses title after scientists spot signs of AI</description>
      <category>watch</category>
      <category>industry</category>
    </item>
    <item>
      <title>TCSP3.0: Template-Based Crystal Structure Prediction with Coefficient-Preserving Element Substitution and Structural</title>
      <link>https://arxiv.org/abs/2610.34029</link>
      <guid isPermaLink="true">https://arxiv.org/abs/2610.34029</guid>
      <pubDate>Sun, 25 Oct 2026 03:34:05 GMT</pubDate>
      <description>Template-based crystal structure prediction (CSP) transfers a known crystal structure to a new composition by element substitution, so its accuracy depends on</description>
      <category>watch</category>
      <category>research</category>
    </item>
    <item>
      <title>When Does GEPA Merge Help? A Mechanism-Level Diagnostic</title>
      <link>https://arxiv.org/abs/2610.34027</link>
      <guid isPermaLink="true">https://arxiv.org/abs/2610.34027</guid>
      <pubDate>Sun, 25 Oct 2026 03:32:01 GMT</pubDate>
      <description>GEPA is among the strongest reflective prompt optimizers for language-model programs, but its system-aware merge operator is unstable across models: merge</description>
      <category>watch</category>
      <category>research</category>
    </item>
    <item>
      <title>The Symmetry of Differential Neural Tangent Kernel under the Orthogonal Group</title>
      <link>https://arxiv.org/abs/2610.34023</link>
      <guid isPermaLink="true">https://arxiv.org/abs/2610.34023</guid>
      <pubDate>Sun, 25 Oct 2026 03:27:08 GMT</pubDate>
      <description>The differential neural tangent kernel (DNTK) is a tool for analyzing the training dynamics of overparameterized physics informed neural networks.</description>
      <category>watch</category>
      <category>research</category>
    </item>
    <item>
      <title>Deep Scientific Reasoning under Physical Constraints: Structure-Aware Spectrum Prediction</title>
      <link>https://arxiv.org/abs/2610.34021</link>
      <guid isPermaLink="true">https://arxiv.org/abs/2610.34021</guid>
      <pubDate>Sun, 25 Oct 2026 03:22:14 GMT</pubDate>
      <description>Structured scientific spectra encode rich physical information while obeying hard constraints, such as conservation laws and characteristic spectral geometry.</description>
      <category>watch</category>
      <category>research</category>
    </item>
    <item>
      <title>BriLLM-MetaPred: A Representation-Prediction Unity Language Model with Adaptive Self-Measurement</title>
      <link>https://arxiv.org/abs/2610.34018</link>
      <guid isPermaLink="true">https://arxiv.org/abs/2610.34018</guid>
      <pubDate>Sun, 25 Oct 2026 03:08:17 GMT</pubDate>
      <description>We present BriLLM-MetaPred, the first engineering implementation of the MetaPred machine-learning framework established by a mathematical theory of AGI</description>
      <category>watch</category>
      <category>research</category>
    </item>
    <item>
      <title>The Information Geometry of Representation Formation</title>
      <link>https://arxiv.org/abs/2610.34015</link>
      <guid isPermaLink="true">https://arxiv.org/abs/2610.34015</guid>
      <pubDate>Sun, 25 Oct 2026 03:01:03 GMT</pubDate>
      <description>Most analyses of neural representations focus on trained endpoints, leaving the process by which internal structure forms comparatively unmeasured.</description>
      <category>watch</category>
      <category>research</category>
    </item>
    <item>
      <title>The Patient Is Always Right? Evaluating Sycophancy in Medical Diagnosis with Graded Patient Beliefs</title>
      <link>https://arxiv.org/abs/2610.34013</link>
      <guid isPermaLink="true">https://arxiv.org/abs/2610.34013</guid>
      <pubDate>Sun, 25 Oct 2026 02:54:42 GMT</pubDate>
      <description>More people are seeking out general-purpose AI models for medical advice, including diagnosis.</description>
      <category>watch</category>
      <category>research</category>
    </item>
    <item>
      <title>Reversible Pseudonymisation of Documents Sent to External LLMs: an Open Bilingual Benchmark and an Engine</title>
      <link>https://arxiv.org/abs/2610.34012</link>
      <guid isPermaLink="true">https://arxiv.org/abs/2610.34012</guid>
      <pubDate>Sun, 25 Oct 2026 02:50:43 GMT</pubDate>
      <description>Professionals send working documents to external large language models.</description>
      <category>watch</category>
      <category>research</category>
    </item>
    <item>
      <title>Cross-User Adaptation for sEMG Gesture Recognition via Continual Learning</title>
      <link>https://arxiv.org/abs/2610.34011</link>
      <guid isPermaLink="true">https://arxiv.org/abs/2610.34011</guid>
      <pubDate>Sun, 25 Oct 2026 02:47:19 GMT</pubDate>
      <description>Prosthetic and assistive interfaces increasingly depend on decoding hand gestures from surface electromyography (sEMG).</description>
      <category>watch</category>
      <category>research</category>
    </item>
    <item>
      <title>Fed-Universe: A Semantic-Geometric-Topological-Human (S-G-T-H) Stack for Negotiated Alignment in Federated Systems</title>
      <link>https://arxiv.org/abs/2610.34009</link>
      <guid isPermaLink="true">https://arxiv.org/abs/2610.34009</guid>
      <pubDate>Sun, 25 Oct 2026 02:43:50 GMT</pubDate>
      <description>Conventional federated learners implicitly optimize for global means, creating a &quot;one-size-fits-all&quot; paradigm that inevitably suppresses the minority under</description>
      <category>watch</category>
      <category>research</category>
    </item>
    <item>
      <title>Partial Modal Tensor Regression</title>
      <link>https://arxiv.org/abs/2610.34008</link>
      <guid isPermaLink="true">https://arxiv.org/abs/2610.34008</guid>
      <pubDate>Sun, 25 Oct 2026 02:42:59 GMT</pubDate>
      <description>With the rapid development of modern scientific research, tensor data have become increasingly prevalent across a wide range of fields.</description>
      <category>watch</category>
      <category>research</category>
    </item>
    <item>
      <title>Steering the Teacher: Improving On-Policy Self-Distillation for Thinking Models</title>
      <link>https://arxiv.org/abs/2610.34005</link>
      <guid isPermaLink="true">https://arxiv.org/abs/2610.34005</guid>
      <pubDate>Sun, 25 Oct 2026 02:34:40 GMT</pubDate>
      <description>On-policy self-distillation (OPSD) trains a student on its own reasoning trajectories using token-level supervision from a reference-conditioned teacher</description>
      <category>watch</category>
      <category>research</category>
    </item>
    <item>
      <title>Resolving the Mobility-Equity-Emissions Trilemma: A Causal Machine Learning Framework for Dynamic Tolling Policy</title>
      <link>https://arxiv.org/abs/2610.33994</link>
      <guid isPermaLink="true">https://arxiv.org/abs/2610.33994</guid>
      <pubDate>Sun, 25 Oct 2026 02:16:15 GMT</pubDate>
      <description>We propose a causal machine learning framework to integrate dynamic tolling for mobility, equity, and climate objectives, resolving a long-standing policy</description>
      <category>watch</category>
      <category>research</category>
    </item>
  </channel>
</rss>
