Vincent Cimino

on the atlas — 28

• Introducing our Science Blog \ Anthropic2 savers
• An overview of areas of control work - by Ryan Greenblatt1 savers
• Humans do acausal coordination all the time — LessWrong1 savers
• [2602.12316] GT-HarmBench: Benchmarking AI Safety Risks Through the Lens of Game Theory1 savers
• models have some pretty funny attractor states — LessWrong2 savers
• AI #155: Welcome to Recursive Self-Improvement1 savers
• 2026 and beyond | Richards Tu's Space2 savers
• Lessons from Moltbook and OpenClaw: The Agentic Internet’s Trust Problem - Irregular1 savers
• [2502.14143] Multi-Agent Risks from Advanced AI1 savers
• Safetywashing — AI Alignment Forum1 savers
• [2408.02565] Reasons to Doubt the Impact of AI Risk Evaluations1 savers
• My response to AI 20271 savers
• We need a better way to evaluate emergent misalignment — LessWrong1 savers
• Emergent Misalignment: Narrow finetuning can produce broadly misaligned LLMs 49 This paper contains model-generated content that might be offensive. 491 savers
• Curius / Onboarding2507 savers
• Dario Amodei — The Adolescence of Technology44 savers
• Research Taste Exercises [rough note] -- colah's blog21 savers
• Vibe physics: The AI grad student \ Anthropic15 savers
• A Guide to Claude Code 2.0 and getting better at using coding agents | sankalp's blog12 savers
• are you high-agency or an NPC? - by Jasmine Sun11 savers
• AI in 2025: gestalt — LessWrong11 savers
• The behavioral selection model for predicting AI motivations — LessWrong10 savers
• CAIS AI Dashboard6 savers
• Workshop Labs PBC5 savers
• Test your interpretability techniques by de-censoring Chinese models — LessWrong5 savers
• How confessions can keep language models honest | OpenAI5 savers
• Why people like your quick bullshit takes better than your high-effort posts — LessWrong2 savers
• Status Is The Game Of The Losers' Bracket — LessWrong2 savers