[KongchangAI]
ConceptFinite-Rollout Variation

Finite Rollout Variance

Instability in voting outcomes caused by sampling only a limited number of responses per prompt, corresponding to sampling variance in statistics.

Timeline (last 90 days)

Nov 9

The paper decomposes the majority-vote supervision error under online direct preference optimization (DPO) into two parts: population-majority bias and prompt-dependent finite-rollout variation.

Unverified50%
Nov 9

Finite-rollout variation refers to the instability of voting results caused by sampling only a limited number of responses per prompt, which in turn affects the quality of pseudo-labels.

Unverified50%
Nov 9

Explicitly handling finite-rollout variation yields consistent performance improvements, especially on sampling-sensitive tasks.

Unverified50%

All Facts (3)

Source Articles