Skip to content

AI 速递 2026-06-10

生成时间:2026/6/10 10:08:06(UTC: 2026-06-10T02:08:06.131Z)

👍 83 · arXiv

Recent AI systems have achieved strong results on a wide range of benchmarks, yet these gains have not translated into economically meaningful deployment across many professional domains. We argue tha…

LatentSkill: From In-Context Textual Skills to In-Weight Latent Skills for LLM Agents

Section titled “LatentSkill: From In-Context Textual Skills to In-Weight Latent Skills for LLM Agents”

👍 51 · arXiv

Agent systems increasingly use textual skills to encode reusable task procedures, but injecting these skills into the prompt at every step incurs substantial context overhead and exposes skill content…

OmniGameArena: A Unified UE5 Benchmark for VLM Game Agents with Improvement Dynamics

Section titled “OmniGameArena: A Unified UE5 Benchmark for VLM Game Agents with Improvement Dynamics”

👍 16 · arXiv

Vision-language model (VLM) agents are increasingly deployed in interactive game environments. Yet game benchmarks for VLM agents typically report a single first-attempt score per (agent, game) pair, …

A Geometric Account of Activation Steering through Angle-Norm Decomposition

Section titled “A Geometric Account of Activation Steering through Angle-Norm Decomposition”

👍 15 · arXiv

Linear activation steering has gained popularity as a simple and empirically effective way to control language model behavior. More recently, spherical steering paradigms have been proposed to address…

SwiftVR: Real-Time One-Step Generative Video Restoration

Section titled “SwiftVR: Real-Time One-Step Generative Video Restoration”

👍 12 · arXiv

Real-time video restoration (VR) for live streams requires high-resolution outputs under strict per-frame latency constraints. Existing one-step diffusion-based VR models remain difficult to deploy on…

  • QQBot now strips model reasoning/thinking scaffolding before native delivery, preventing raw <thinking> content from leaking into channel replies. (#89913, #90132) Thanks @openper…

链接https://github.com/openclaw/openclaw/releases/tag/v2026.6.5

Changes since langchain==1.3.5

release(langchain): 1.3.6 (#38001) fix(langchain): preserve summarization trigger compatibility (#38000)…

链接https://github.com/langchain-ai/langchain/releases/tag/langchain%3D%3D1.3.6

  • Migrate @listen/@router runtime to read from FlowDefinition
  • Add pluggable default backends for memory, knowledge, rag, and flow
  • Update changelo…

链接https://github.com/crewAIInc/crewAI/releases/tag/1.14.7a4

  • Code mode can now call standalone web search directly, including from nested JavaScript tool calls, and receive plaintext search results. (#26719)
  • Tool and connector input schemas …

链接https://github.com/openai/codex/releases/tag/rust-v0.139.0

AI misidentification results in wrongful arrest; man seeks justice

Section titled “AI misidentification results in wrongful arrest; man seeks justice”

Article URL: https://www.wsoctv.com/news/local/ai-misidentification-results-wrongful-arrest-man-seeks-justice/I7UQJWV33FBN3LMKHCSXI6FIVA/ Comments URL: https://news.ycombinator.com/item?id=48468789 Points: 75

来源Hacker News AI

If Claude Fable stops helping you, you’ll never know

Section titled “If Claude Fable stops helping you, you’ll never know”

Article URL: https://jonready.com/blog/posts/claude-fable5-is-allowed-to-sabotage-your-app-if-youre-a-competitor.html Comments URL: https://news.ycombinator.com/item?id=48467896 Points: 504

来源Hacker News AI

Apple’s AI Can Now Change Your Passwords. What Could Possibly Go Wrong?

Section titled “Apple’s AI Can Now Change Your Passwords. What Could Possibly Go Wrong?”

Article URL: https://www.kylereddoch.me/blog/apples-ai-can-now-change-your-passwords-what-could-possibly-go-wrong/ Comments URL: https://news.ycombinator.com/item?id=48465744 Points: 78

来源Hacker News AI

CEOs who think AI replaces their employees are just bad CEOs

Section titled “CEOs who think AI replaces their employees are just bad CEOs”

Article URL: https://www.techdirt.com/2026/06/09/ceos-who-think-ai-replaces-their-employees-are-just-bad-ceos/ Comments URL: https://news.ycombinator.com/item?id=48465675 Points: 435

来源Hacker News AI

Article URL: https://naokishibuya.github.io/blog/2022-12-30-gpt-2-2019/ Comments URL: https://news.ycombinator.com/item?id=48465269 Points: 255

来源Hacker News AI

Article URL: https://www.apollo.com/wealth/the-daily-spark/where-is-the-ai-jobs-crisis Comments URL: https://news.ycombinator.com/item?id=48464333 Points: 137

来源Hacker News AI

System Card: Claude Fable 5 and Claude Mythos 5 [pdf]

Section titled “System Card: Claude Fable 5 and Claude Mythos 5 [pdf]”

Article URL: https://www-cdn.anthropic.com/d00db56fa754a1b115b6dd7cb2e3c342ee809620.pdf Comments URL: https://news.ycombinator.com/item?id=48463811 Points: 211

来源Hacker News AI

System Card [pdf]: https://www-cdn.anthropic.com/d00db56fa754a1b115b6dd7cb2e3c3

Comments URL: https://news.ycombinator.com/item?id=48463808 Points: 1807

来源Hacker News AI