Built Among Us Sim to Test LLM Deception: Claude vs. Kimi
panickssery · x · 2026-08-22
Inspired by the Generative Agents Simulacra paper, the author built simulations to observe how LLMs play games against each other.
- Testbed: Chose "Among Us" (Mafia style game) and mapped every pixel to build a custom sim (pinned on profile).
- Matchup: Had Claude play against Kimi to gather data on LLM deception in game scenarios.
- Duration: Working on variations of this idea since October 2025 (10 months).
Related event: Among Us-Based Environment Tests LLM Deception(4 posts)→
More from coding & agent
- Fable 5 vs Ox Alpha spaceship demo: Anthropic's model still dominates game coding — ChrisGPT · 2026-08-22
- Agent Security Guide: Least Privilege to Prevent Cascading Failures — blaizedsouza · 2026-08-22
- OpenCode Senses adds local vision to text-only coding models — JeremyCMorgan · 2026-08-22
- FreeToken: Run 290B+ frontier MoE models locally at interactive speeds — SysPsych · 2026-08-22
- Coco: Local Voice Assistant Powered by Inkling with Proactive Collaboration — EchoShao8899 · 2026-08-22
- Claude oneshots servo motor config with hardware specs — _Stocko_ · 2026-08-22