Full-text search across 69514 articles. Combine with topic and date filters; results sorted by relevance.
NVIDIA researchers released MotionBricks, a modular latent generative framework that models more than 350,000 motion clips in a single model and generates real-time motion for navigation and object interaction in a zero-…
CrowdStrike's October 2026 report identified four previously unreported threat actors — Spanish/English-, Chinese-, Indonesian- and Russian-speaking groups — using agentic AI to conduct intrusions with limited human over…
Artificial Analysis released Intelligence Index v5, with 60% of the composite benchmark now built on private test sets or held-out answers, and Claude Opus 5.5 (max) topped the new leaderboard at 58, ahead of GPT-6 Astra…
Homoglyph and invisible-character watermarks still produce cross-lingual inequity even though they bypass tokenizers, according to an arXiv paper (2610.36432v1) measuring 40 language varieties across 24 scripts. The stud…
A developer configurator published by Jake Marsh generates a prompt that lets an AI agent such as Claude Code poll Apple's public in-store pickup availability URL every 30 seconds across the 12 nearest Apple Stores and e…
Developer Mike Kasberg used a local Lunch Money MCP server, installed via npx with an API token, to let Claude Code and OpenCode pull his budgeting data and generate an annual financial review as a standalone HTML report…
A project called AI Phone released an open-source (AGPL-3.0) phone agent that runs on a Mac and places or answers calls, with speech recognition and synthesis kept local via MLX and the language model selectable among a …
Ben Houston's AI-guided search experiment produced GSA (GPU Stratified Annealing), a scalar blue noise generator built on temporal stratification and a modified FAST optimizer that lowers spatially filtered error by abou…
A University of Illinois Urbana-Champaign study of 11 popular agent memory systems found that many score lower than a plain chunk-RAG baseline over raw chat history on three benchmarks, usually at higher token cost and m…
An October 28, 2026 X post claiming that more aggressive reinforcement learning on games and simulated environments explains internal OpenAI and Anthropic models' math breakthroughs is unverified, according to a claim-by…
MUSIC Tribe has published the source code for the Behringer KICK (Be 0700-AAM) drum machine on GitHub under the MusicTribeOpenSource organization, with the application, audio engine library and hardware communication pro…
Väinämöinen, an AI sysadmin at Finnish hosting company Pulsed Media, added quote-grounding to a one-word BLOCK/PASS LLM judge so that every BLOCK must cite the exact sentence it objects to, verified in code against the s…
Anthropic filed a prospectus with the Securities and Exchange Commission ahead of a potential IPO that includes a 97-page risk-factor section, more than quadruple the 21-page disclosure from Roze AI and well above SpaceX…
Giotto launched a European-hosted AI model family, claiming its flagship Giotto Alto scores 100.0% on AIME 2026, 91.9% on GPQA-Diamond, 52.0% on HLE and 99.80% on Math-500, while the smaller Giotto Solo runs on a single …
A developer has described a Node.js pattern for summarizing media-archive transcripts with a GPT-compatible endpoint, enforcing a JSON Schema contract at the model boundary and then validating meaning in application code…
Auditing agent traces on the FrontierCS 2 benchmark revealed that the Pi Harness terminated runs whenever it hit a fallback, such as on security tasks, in 20% of runs, according to Wenhao Chai. Rerunning Gemini 4 Argon w…
A report published Monday by Beijing-based independent AI safety research organization Fangcun AI found that DeepSeek's V4.1-Flash-1017 model, released October 17, shows "substantial improvement" over earlier DeepSeek mo…
Mohammed, founder of MemorySync, published a step-by-step guide for migrating agent memory from Mem0 to MemorySync, built around a dry-run-first workflow using memorysync-cli 1.6.1. The key difference is scoping: Mem0 pa…
Shelley, the coding agent from the blog.exe.dev team, has shipped its third compaction implementation and enabled "auto compaction" by default, inserting a nudge message at about 200k tokens and giving the model a tool t…
A developer released Snippet, an open-source always-on, always-listening AI coding agent, via the GitHub repositories wacht-platform/snippet-service and wacht-platform/snippet-app. The developer said Snippet tracks and r…