Unverified50% confidenceFactExact time
MCPO maps the margin to a score via two fixed thresholds and, while keeping the mean total advantage of each solution unchanged, distributes the advantage across tokens.
1
Sources
50%
Confidence
Long-term
Relevance
11/11/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedMCPO uses the top-2 margin (the probability difference between the highest-probability and second-highest-probability tokens) of the policy's next-token distribution to measure the model's degree of hesitation, with a smaller margin indicating greater model uncertainty.70% similarUnverifiedA more reliable way to use thresholds is to combine relative ranking (such as keeping passages in the top K results whose scores are significantly higher than the rest) with verification-based judgment (having the model judge whether a passage truly supports the answer), rather than relying on a single static threshold.64% similarUnverifiedWhen the reward density vanishes at the support endpoints, for 1 ≤ p ≤ 2, the resolution-independent root-n rate is still maintained as long as the reward quantile function satisfies continuity and endpoint monotonicity.64% similarUnverifiedIn multi-objective optimization, the Pareto frontier refers to a set of mutually non-dominated solutions; in the deviation-support trade-off, each point represents the optimal trade-off of maximum deviation for a given support or minimum support for a given deviation.63% similarUnverifiedCompressing token count and improving instruction quality are two orthogonal goals that need to be addressed separately.63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1157950API
curl https://kongchang.com/api/v1/knowledge/claims/1157950MCP
get_claim(id=1157950)