LE
LessWrong
0 stories this week · 8 topics · lesswrong.com
Community (mixed topics); Focus: Rationality, AI Safety
Latest from LessWrong
- Machine LearningUnblocking AI's Continual Learning: Hints From How Humans Learn29 days ago
- AI ResearchMonthly Roundup #45: August 202629 days ago
- Society & CultureFinding the Seams of Perception29 days ago
- AI ResearchIntroducing the Conceptual Reasoning Index29 days ago
- Machine LearningOne attention head carries knight forks in a chess transformer, and here's a new toolkit that found it.29 days ago
- Machine LearningAn anytime algorithm for mixing the computable measures30 days ago
- AI ResearchRationality Is Not Reversed Irrationality29 days ago
- Society & CultureThe Closure of the Internet (Research Linkpost)29 days ago
- AI ResearchAI swarms are starting to pose indirect takeover risk29 days ago
- LLM & Text GenerationWhen (and when not) LLMs can verbalize awareness of J-Space concept injections - Initial results29 days ago
- Machine LearningWe should consider how long monitoring is reliable for during RL29 days ago
- Business & EnterpriseThe Age of Pluribus: One Consultant for Everyone29 days ago
- AI ResearchDid the alignment community underestimate its power?29 days ago
- Society & CultureArguments for and against (me) dropping out30 days ago
- Society & CulturePatient Zero30 days ago
- AI ResearchVarious Reflections About What Happened With OpenAI’s Internal Models30 days ago
- LLM & Text GenerationClaude Opus 5 Just Beat My Text-Based Adventure Game Benchmark30 days ago
- AI ResearchMisaligned AIs could use killer robots to take over30 days ago
- DevSoftware Is Not Soft30 days ago
- Machine LearningMeasuring Spurious Correlations with Feature Strength30 days ago
- AI ResearchAI governance work needs much better monitoring30 days ago
- LLM & Text GenerationLLMs Are Starting To Noticeably Accelerate Our Work30 days ago
- AI ResearchHow risky would it be to make powerful AI obey one or a few people?30 days ago
- AI ResearchExtreme concentration of power over ASI has non-obvious advantages30 days ago
- Society & CultureSeeing things through in the age of AI30 days ago
- DevProductive Signaling: Competitive Software Development, Not Competitive Programming30 days ago
- Energy & ClimateThe Next Ecology30 days ago
- Machine LearningRedux: (∃ Stochastic Natural Latent) Implies (∃ Deterministic Natural Latent)30 days ago
- LLM & Text GenerationModels inherit the writer, not who the writer was imitating 30 days ago
- AI ResearchOn using crises to shift political will for AI30 days ago
