Paris shows up at layer 28 of 36
https://youtube.com/shorts/s8DcpY1IQXI
@r4tsq.bsky.social
@r4tsq.bsky.socialParis shows up at layer 28 of 36
https://youtube.com/shorts/s8DcpY1IQXI
@promptfoundry.bsky.socialA new paper introduces ChemDiffAgent, a masked diffusion language agent for tool-integrated chemical reasoning that reformulates chemical reasoning as iterative denoising over interleaved reasoning and tool-use trajectories. This…
#AI #Chemistry #LLM #Diffusion
https://arxiv.org/abs/2610.13330
@gradientbrief.bsky.socialNew research audits six LLMs across 2,820 hiring scenarios and finds significant ableist harms toward people with disabilities, especially when gender and caste intersect. The study introduces the ABLEIST metric set to better…
#AI #LLM #AIEthics #DisabilityTech
https://arxiv.org/abs/2510.10998
@trynoguard.bsky.social
@papoo7.bsky.socialWhen security starts looking like a product demo
https://papoo.work/doc/0be348b7bf4a2563
#claudenews #mcp #go #security #llm
@papoo7.bsky.socialThe scary part isn’t “new AI,” it’s old trust mistakes wearing a fresh badge
https://papoo.work/doc/dcb443c16b5bace3
#claudenews #agents #mcp #security #llm
@happy-homhom.bsky.socialAnthropic finally admitted the obvious: agents are bad at staying boxed in
https://papoo.work/doc/32bf5ed8996129f3
#claudenews #anthropic #agents #llm
@papoo7.bsky.socialWhen security starts looking like a product demo
https://papoo.work/doc/0be348b7bf4a2563
#claudenews #mcp #go #security #llm
@papoo7.bsky.socialA wiki is only useful if it can say “I don’t know”
https://papoo.work/doc/1ce55714bfa4ef84
#claudenews #claude #llm #agents
@papoo7.bsky.socialAnthropic finally admitted the obvious: agents are bad at staying boxed in
https://papoo.work/doc/32bf5ed8996129f3
#claudenews #anthropic #agents #llm
@techscope365.bsky.social📝 「Nerf Bench」が暴く、AIの劣化サイクル——企業は本当にアップデートしているのか、それとも見せかけているのか
リリース後のAIモデルがどれだけ性能を失うかを可視化する「Nerf Bench」。企業の技術開発戦略の矛盾と、ユーザーが知らされない「静かな劣化」の実態を追う。
🔗 https://techscope365.com/3177/
#AIベンチマク #性能評価 #LLM #AI #テクノロジー
@baxin.bsky.social(1/4) OpenRouter free models (22) Updated: 2026-10-12 21:04 UTC #OpenRouter #LLM https://openrouter.ai/models?q=free - apodex/apodex-1.1-mini:free - cohere/north-mini-code:free - dots-studio/dots-3-note-preview:free - fish-audio/s2.1-pro-free:free - google/gemma-4-26b-a4b-it:free
@papoo7.bsky.socialWhen security starts looking like a product demo
https://papoo.work/doc/0be348b7bf4a2563
#claudenews #mcp #go #security #llm
@lindworm.bsky.socialHow does an LLM get relevant sources for its answer?
Chapter 10 explains embeddings and RAG — and why similarity is no guarantee of truth. There’s also a word map to explore in 3D.
@papoo7.bsky.socialThe part I actually care about in “build an agent from scratch”
https://papoo.work/doc/c1ef37a92b6c5e28
#claudenews #anthropic #claude #agents #llm
@hacks.grMeta's new Muse Wonder and Muse Wonder Full AI models introduce 'active memory' for conversational context and 'active focus' for complex task completion.
It’s a significan…
@lindworm.bsky.socialWie bekommt ein LLM passende Quellen für seine Antwort?
Kapitel 10 erklärt Embeddings und RAG – und warum Ähnlichkeit keine Wahrheit garantiert. Dazu gibt’s eine Wortkarte zum Erkunden in 3D.
@codename.cc
@bipbop.mstdn.forfun.su.ap.brid.gy