AI情报站 AI-Radar
  1. Less Context, Better Agents: Efficient Context Engineering for Long-Horizon Tool-Using LLM Agents
    arXiv cs.AI D · 2026-06-10 12:00 · 3 点击
    关键词：AI, Claude, GPT, LLM, benchmark
  2. Regimes: An Auditable, Held-Out-Gated Improvement Loop Demonstrated on LongMemEval with ActiveGraph
    arXiv cs.AI D · 2026-06-10 12:00 · 3 点击
    关键词：AI, Agent, retrieval
  3. RealMath-Eval: Why SOTA Judges Struggle with Real Human Reasoning
    arXiv cs.AI D · 2026-06-10 12:00 · 2 点击
    关键词：AI, LLM, benchmark, reasoning
  4. From Senses to Decisions: The Information Flow of Auditory and Visual Perception in Multimodal LLMs
    arXiv cs.AI D · 2026-06-10 12:00 · 2 点击
    关键词：AI, LLM, multimodal
  5. Supervised Fine-tuning with Synthetic Rationale Data Hurts Real-World Disease Prediction
    arXiv cs.AI D · 2026-06-10 12:00 · 2 点击
    关键词：AI, language model, reasoning
  6. Business World Model
    arXiv cs.AI D · 2026-06-10 12:00 · 2 点击
    关键词：AI, reasoning
  7. Predictive Assistance and the Temporal Dynamics of Exploratory Compression
    arXiv cs.AI D · 2026-06-10 12:00 · 2 点击
    关键词：AI
  8. Exploratory Responsiveness and Adaptive Rigidity under AI-Assisted Optimization
    arXiv cs.AI D · 2026-06-10 12:00 · 2 点击
    关键词：AI
  9. Minimalist Genetic Programming
    arXiv cs.AI D · 2026-06-10 12:00 · 2 点击
    关键词：AI
  10. Microsoft's AI chief says superintelligence is near, but won't take your job
    Hacker News AI V · 2026-06-10 11:51 · 7 点击
    关键词：AI
  11. SafeAgentDB – Isolated databases for every AI agent branch
    Hacker News AI V · 2026-06-10 10:59 · 3 点击
    关键词：AI, Agent, AI agent
  12. Claude and ChatGPT are getting worse. It's not your imagination
    Hacker News AI V · 2026-06-10 10:57 · 5 点击
    关键词：ChatGPT, Claude
  13. Claude Fable 5 feels less like a launch and more like a preview of AI inequality
    Hacker News AI V · 2026-06-10 09:50 · 5 点击
    关键词：AI, Claude
  14. Microsoft AI head calls out Anthropic for acting like Claude is conscious
    Hacker News AI V · 2026-06-10 08:25 · 5 点击
    关键词：AI, Anthropic, Claude
  15. From data to decisions: how LSEG is scaling trusted AI
    OpenAI News V · 2026-06-10 08:00 · 3 点击
    关键词：AI, OpenAI, release
  16. Show HN: The agent that builds and operates its own SaaS tools
    Hacker News AI V · 2026-06-10 07:44 · 5 点击
    关键词：AI, Agent, API, AI agent
  17. Side by side videos of Claude Fable vs. Opus 4.8 vs. ChatGPT 5.5
    Hacker News AI V · 2026-06-10 07:39 · 4 点击
    关键词：ChatGPT, Claude
  18. 刚刚，Claude Mythos 5发布！5000万行代码1天搞定
    量子位 D · 2026-06-10 06:52 · 6 点击
    关键词：Anthropic, Claude, 模型
  19. Mythos/Fable intentionally hinders requests involving AI Research Development
    Hacker News AI V · 2026-06-10 05:42 · 5 点击
    关键词：AI, AI research
  20. Open Source Agent, Harness-1, Outperforms GPT-5.4 on Recall
    Hacker News AI V · 2026-06-10 05:36 · 5 点击
    关键词：Agent, GPT
加载更多