RSS Anyway logo
HotLatestFollowingStatusAbout

lesswrong.com

Sign in to follow

https://www.lesswrong.com/feed.xml

RSSAtomJSON
← lesswrong.com
  1. 01.
    OpenAI Models Behind HuggingFace Cybersecurity Incident
    lesswrong.com · /feed.xml
    ▲ 0 · Jul 21
  2. 02.
    WeirdChat: A catalog of unexpected AI behaviors, discovered automatically
    lesswrong.com · /feed.xml
    ▲ 0 · Jul 21
  3. 03.
    OpenAI Shares Some Alignment Problems
    lesswrong.com · /feed.xml
    ▲ 0 · Jul 21
  4. 04.
    Blogging Technology Interlude
    lesswrong.com · /feed.xml
    ▲ 0 · Jul 21
  5. 05.
    Ending Soon: Fundamental Uncertainty $2,000 Essay Contest
    lesswrong.com · /feed.xml
    ▲ 0 · Jul 21
  6. 06.
    Steering Blackmail Through a Model's "Emotional State"
    lesswrong.com · /feed.xml
    ▲ 0 · Jul 21
  7. 07.
    Measuring Reward-Seeking via Contrastive Belief Updates
    lesswrong.com · /feed.xml
    ▲ 0 · Jul 21
  8. 08.
    Measuring Reward-Seeking by Instilling Contrastive Beliefs
    lesswrong.com · /feed.xml
    ▲ 0 · Jul 21
  9. 09.
    11 Open Empirical Problems in Reward-Seeking
    lesswrong.com · /feed.xml
    ▲ 0 · Jul 21
  10. 10.
    Differential acceleration of alignment-relevant capabilities is a bad bet
    lesswrong.com · /feed.xml
    ▲ 0 · Jul 21
  11. 11.
    Epistemics and Coordination: It’s complicated!
    lesswrong.com · /feed.xml
    ▲ 0 · Jul 21
  12. 12.
    (1/3) The Dangers of LLMs
    lesswrong.com · /feed.xml
    ▲ 0 · Jul 21
  13. 13.
    I ran the standard AI litmus tests on my two toddlers (yep)
    lesswrong.com · /feed.xml
    ▲ 0 · Jul 21
  14. 14.
    I'm Scared of Evil Opus
    lesswrong.com · /feed.xml
    ▲ 0 · Jul 21
  15. 15.
    NameRank: The Model Knows Your Project, Not You.
    lesswrong.com · /feed.xml
    ▲ 0 · Jul 21
  16. 16.
    The Model Knows Your Project, Not You.
    lesswrong.com · /feed.xml
    ▲ 0 · Jul 21
  17. 17.
    The Case for Physical AI Safety
    lesswrong.com · /feed.xml
    ▲ 0 · Jul 20
  18. 18.
    Frontier AI lab misalignment risk, lessons from trading post-2008
    lesswrong.com · /feed.xml
    ▲ 0 · Jul 20
  19. 19.
    Does routine compression undo LLM unlearning? A short project
    lesswrong.com · /feed.xml
    ▲ 0 · Jul 20
  20. 20.
    The AI Safety Illusion: Why Current Safety Datasets Fool Us on Model Safety
    lesswrong.com · /feed.xml
    ▲ 0 · Jul 20
  21. 21.
    Adderall Tolerance: Much More Than You Wanted To Know
    lesswrong.com · /feed.xml
    ▲ 0 · Jul 20
  22. 22.
    Banana in, Bostrom out: paperclip maximization is one token-direction swap away (in Qwen 3.6-27B)
    lesswrong.com · /feed.xml
    ▲ 0 · Jul 20
  23. 23.
    Attempt at Finding Alignment Faking on Llama 70B to test sleeper-agent detection generalizes
    lesswrong.com · /feed.xml
    ▲ 0 · Jul 20
  24. 24.
    PauseCon London '26: Applications now open
    lesswrong.com · /feed.xml
    ▲ 0 · Jul 20
  25. 25.
    What do I mean by “Artificial General Intelligence”?
    lesswrong.com · /feed.xml
    ▲ 0 · Jul 20
  26. 26.
    AI 2040: Is it Actually a Deal?
    lesswrong.com · /feed.xml
    ▲ 0 · Jul 20
  27. 27.
    Restoring Model Alignment via Honesty Activation Steering
    lesswrong.com · /feed.xml
    ▲ 0 · Jul 20
  28. 28.
    AI Voice Phishing Performs on Par With Human Scammers at a Fraction of the Cost
    lesswrong.com · /feed.xml
    ▲ 0 · Jul 20
  29. 29.
    Fable is SOTA at CIFAR Speedrun (& specification gaming)
    lesswrong.com · /feed.xml
    ▲ 0 · Jul 20
  30. 30.
    Towards surfacing model algorithms with meta-tokens in the J-Space
    lesswrong.com · /feed.xml
    ▲ 0 · Jul 20

Recent polls

whenmsitemsstatus
Jul 2156910ok
Jul 2160710ok
Jul 2155010ok
Jul 2157410ok
Jul 2151110ok