Abstract
LLM-as-judge evaluation reduces response quality assessment to a single holistic A/B preference choice, providing no mechanism to isolate which quality dimensions drove the preference or distinguish model errors from genuine label ambiguity. We propose SESSE (Sketch, Expand, Sort, Summarize, Evaluate), a training-free framework that decomposes holistic judgment into structured sub-questions mined directly from the judge's own error cases; requiring no oracle responses, task-specific rubrics, or fine-tuning. On RewardBench (n=1,000), SESSE achieves near-parity with the chain-of-thought baseline and is competitive with RISE-Judge-32B (92.7%), a fine-tuned specialist, while remaining fully training-free. Per-criterion vote evidence provides an interpretable audit trail for diagnosing label ambiguity and judge failure modes unavailable from a single holistic output token.
Abstract (translated)
URL
https://arxiv.org/abs/2608.18303
