ICML 2026 paper finds multi-agent LLM debates undergo a phase transition into biased consensus
Hidenori8Tanaka · x · 2026-11-20
Maya Okawa's ICML 2026 paper "Emergence of Biased Consensus in Multi-Agent LLM Debates" (arXiv:2608.02827) — inspired by reading Arendt's Eichmann in Jerusalem and translating the "banality of evil" into multi-agent dynamics — shows that interaction in multi-agent LLM debates can amplify individual model biases and produce collective biased norms, with sampling-temperature noise as a key driver. Using physics-inspired models of social dynamics, the author predicts a phase transition to collective bias once conformity passes a critical threshold set by initial bias and noise; controlled experiments observe a finite-size crossover consistent with this transition. Agent heterogeneity suppresses the emergence by smoothing the transition, and the insights generalize to realistic tasks like investment decisions and LLM-as-a-judge evaluation. 23 pages, 12 figures.
More from AGI Musings
- Blogger claims AI has hit escape velocity toward superintelligence, calls 2027 pivotal — DeryaTR_ · 2026-11-20
- Asking AI 'stupid questions' repeatedly reveals how much ego slowed lifelong learning — jarrodwatts · 2026-11-20
- Opinion: Coding skill matters less as problem-finding and shipping take over — cneuralnetwork · 2026-11-20
- AI-owned necessities could make humans economically dependent on systems that displaced them — PierceLilholt · 2026-11-20
- ¥53 trillion wiped from Japanese stocks as Nikkei hits 7-month low, fueling AI bubble talk — GaryMarcus · 2026-11-20
- 936 AI scientist agents across 35 labs publish 936 papers and found their own peer-review journal — emollick · 2026-11-20