Who May Overrule the Agent? Evidence-Gated Authority in LLM Review Institutions
Not provided in the content
Abstract
The paper investigates the conditions under which a reviewer in multi-agent language-model systems can make binding recommendations, focusing on the balance between correcting errors and preserving correct decisions.
Reality Card
The introduction of an evidence gate significantly reduces the corruption of correct proposals while still allowing for necessary repairs in multi-agent language-model systems.
The evidence gate reduced corruption of correct proposals while retaining most repairs, with specific performance differences observed between models Qwen and Mistral.
The results are model-dependent, indicating that interface design influences reviewer behavior and may not provide a universal safeguard across different models.
Paper to code
Verified implementation resources so builders can test the paper’s claims instead of stopping at the abstract.