Code-MUE estimates code-LLM uncertainty by executing sampled programs on synthetic inputs and computing the Von Neumann entropy of a behavioral-agreement graph, reporting Spearman correlations as strong as -0.98 with functional correctness.
Title resolution pending
1 Pith paper cite this work, alongside 1 external citations. Polarity classification is still indexing.
1
Pith paper citing it
1
external citations · OpenAlex
fields
cs.SE 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Code-MUE: Measuring Code LLMs' Uncertainty through Execution-based Semantic Interaction Graphs
Code-MUE estimates code-LLM uncertainty by executing sampled programs on synthetic inputs and computing the Von Neumann entropy of a behavioral-agreement graph, reporting Spearman correlations as strong as -0.98 with functional correctness.