CT-Bench, a benchmark for LLM agent cooperation via self-negotiated contracts, accepted at COLM 2026
maxhkw · x · 2026-11-21
The paper "Commitment to Cooperation with Self-Negotiated Contracts" is accepted at COLM 2026. The team introduces CT-Bench, an environment for studying cooperation, bargaining, and commitment between LLM agents: agents negotiate payment arrangements, commit to contracts, then move through move, contract-execution, and payment-execution phases (counterpart can accept or reject payment) to test whether self-negotiated contracts sustain cooperation.
More from coding & agent
- Google's Pro² Agent hits 76% medal rate on MLE-Bench, a new record — dair_ai · 2026-11-21
- Open-source PlaceBench tests 6 voice assistants on 150 real-place questions: Dots 92, Ask Maps 49 — whowasntright · 2026-11-21
- App plus new database cluster deployed in 12 seconds with Cloud Run --no-build and AlloyDB Express — rseroter · 2026-11-21
- Claude Code 2.1.330 ships with 142 CLI changes and confirm-before-risky-actions policy — ClaudeCodeLog · 2026-11-21
- Claude builds Miami-style town with 2,538 Blender assets, rendered in Unreal in 74 minutes — piotrbinkowski · 2026-11-21
- Antigravity harness restrictions spark ban-risk concerns for Gemini subscribers — jeremyphoward · 2026-11-21