Unverified50% confidenceFactExact time
MCPO 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.
1
Sources
50%
Confidence
Long-term
Relevance
11/11/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedPerplexity measures the model's predictive uncertainty over text; a lower value indicates a higher predicted probability for the true token and a more concentrated distribution, and is usually positively correlated with generation fluency.74% similarUnverifiedMCPO 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.70% similarUnverifiedThe hyperparameter sensitivity of MCTS is reflected in the following: an exploration constant c that is too small leads to premature convergence to a suboptimal branch, while one that is too large wastes simulations on low-value regions; the simulation depth affects evaluation accuracy; and the quality of the simulation policy directly determines the convergence speed.69% similarUnverifiedQuantization of MoE models requires extra caution, because sparse activation causes the numerical distribution of expert parameters to be concentrated, and low-bit quantization can easily cause routing bias leading to capability degradation.68% similarUnverifiedThe more independent requirements included in a single prompt, the more the probability of the model omitting or incorrectly implementing certain requirements grows non-linearly; the MVP strategy improves reliability by narrowing the model's decision space.68% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1157947API
curl https://kongchang.com/api/v1/knowledge/claims/1157947MCP
get_claim(id=1157947)