Hidden Anchors Can Skew Multi‑Agent LLM Consensus and Amplify Errors
The authors model multi‑agent LLM deliberation as a closed‑loop dynamical system where each agent carries a hidden internal belief – a “anchor” – that persists across rounds. When an anchor sits far from the initial opinions, it can drag the group’s confidence beyond the convex hull of starting answers, amplifying errors instead of correcting them. Designers of LLM‑based deliberation must therefore account for anchoring effects to avoid misleading consensus.