Hacker News

New stories

Live mirror
30 storiesupdated just nowView source snapshot
  1. AOS GLM Language Engine – Beta Test(github.com/thepikey)
    1 pointsby 2 minutes ago
    —discuss
  2. Ask HN: How do you test Lambda/ECS code before deploying?
    1 pointsby 4 minutes ago
    2comments
  3. Work, Wellbeing, and Choice: Empirical Lessons for AI Futures(arxiv.org)
    1 pointsby 4 minutes ago
    —discuss
  4. Local AI News Feed(local-ai-signal.vercel.app)
    1 pointsby 6 minutes ago
    1comments
  5. The Fine-Grained City Needs an Operating System(courtyardurbanist.com)
    1 pointsby 9 minutes ago
    —discuss
  6. Small Modular Reactors: a short history of a nuclear pipe dream(engelsbergideas.com)
    1 pointsby 10 minutes ago
    —discuss
  7. LensVLM: Compressing long context as images, expanding only relevant pages(huggingface.co)
    2 pointsby 10 minutes ago
    —discuss
  8. Ask HN: Do you offload your coding to AI, or keep your skills sharp?
    3 pointsby 11 minutes ago
    —discuss
  9. Can internal model transparency tame the AI race?(karthiktadepalli.com)
    1 pointsby 12 minutes ago
    —discuss
  10. The meta-harness for coding agents(soloterm.com)
    1 pointsby 12 minutes ago
    —discuss
  11. Show HN: Jev-mice – mouse colony simulation using Jev and deterministic engine(carsonsweet.com)
    1 pointsby 13 minutes ago
    —discuss
  12. Discover Solo: The Ultimate AI-Integrated Control Panel(master.dev)
    1 pointsby 16 minutes ago
    —discuss
  13. Steam Frame Teardown(ifixit.com)
    1 pointsby 17 minutes ago
    —discuss
  14. SpaceX Starshield(wikipedia.org)
    1 pointsby 18 minutes ago
    —discuss
  15. Mexican Navy's tall ship Cuauhtémoc will be at the Embarcadero all week(kqed.org)
    1 pointsby 18 minutes ago
    —discuss
  16. A Worthwhile Trade(nejm.org)
    2 pointsby 19 minutes ago
    —discuss
  17. Loopjacking in A2A Implementations: Hijacking Human-in-the-Loop Approvals(adithyanak.com)
    1 pointsby 23 minutes ago
    —discuss
  18. How AI Helps and Hampers Junior Workers(nominalnews.com)
    2 pointsby 24 minutes ago
    —discuss
  19. Jackrong/Qwopus3.8-27B-Flash-V2(huggingface.co)
    1 pointsby 24 minutes ago
    —discuss
  20. Nemotron 3 Diarization(huggingface.co)
    2 pointsby 26 minutes ago
    —discuss
  21. CKKS – Encryption and Decryption(jeremykun.com)
    1 pointsby 26 minutes ago
    —discuss
  22. What AI pays the people who train it(dpdns.org)
    2 pointsby 27 minutes ago
    —discuss
  23. Cloud Agents Are Inevitable AI Prisons(normanponte.io)
    2 pointsby 30 minutes ago
    —discuss
  24. Ask HN: Why so many dead comments lately?
    2 pointsby 31 minutes ago
    2comments
  25. Illinois, California governors fast-track AI safety measures with EOs(transparencycoalition.ai)
    2 pointsby 32 minutes ago
    —discuss
  26. Gvnr(gvnr.io)
    5 pointsby 33 minutes ago
    —discuss
  27. HySparse2: Hybrid Sparse Attention with Two-Level KV Sharing(arxiv.org)
    1 pointsby 35 minutes ago
    —discuss
  28. The Vigilante Who's Taking Down Recruitment Scammers, One Fake Job Ad at a Time(wsj.com)
    3 pointsby 35 minutes ago
    1comments
  29. Pretraining data, not verifiability, is why LLMs are good at math (and coding)(lesswrong.com)
    1 pointsby 36 minutes ago
    —discuss
  30. Microsoft Publisher will no longer be supported after October 2026(support.microsoft.com)
    3 pointsby 36 minutes ago
    2comments

Repligraphs: A Primitive for AI Provenance

3 pointsby 55m agolukechampine.com
Open original View on HN
1 comments
17m agoHN ↗

How does this work since LLM outputs aren't deterministic? Most users don't have full control of their LLMs.

Open thread on HN ↗