Model Releases
Valid Stopping in Adaptive Generator-Verifier Loops
arXiv:2610.06432v1 Announce Type: cross Abstract: Numerous agentic workflows are based on a generator-verifier loop: a generator proposes candidates, a cheap verifier scores them, and the workflow ter
arXiv:2610.06432v1 Announce Type: cross Abstract: Numerous agentic workflows are based on a generator-verifier loop: a generator proposes candidates, a cheap verifier scores them, and the workflow terminates when a proposal is verified as good enough. The verifier typically proxies a more costly ground-truth oracle, and as the generator searches adaptively against it, false acceptances may accumulate. Proposals can pass the proxy but fail under the costlier ground-truth check. We study when to stop these loops while controlling the false discovery rate of the accepted proposals. Our construction introduces tools of independent interest in distribution-free statistical testing and conformal risk control, including analysis of e-values constructed through index betting and a novel conformal risk control procedure for non-monotone losses. We validate the approach in synthetic settings and on a protein-design benchmark.
Related
- Don't Read the Log: Execution Traces Contaminate Verifiers in Video-Generation Agents
- AdaSpark: Adaptive DSpark with Online Learning for Tree Verification and N-gram Fill
- PPO guided Agentic Pipeline for Adaptive Prompt Selection and Test Case Generation
Source: arXiv cs.LG | 2026-10-06