Multi-sample UQ methods like Semantic Entropy show no clear advantage over single-sample methods for LLM function-calling, but both improve when adapted using abstract syntax tree clustering for multi-sample approaches and semantic token selection for logit-based single-sample scores.
, yJ }inserted here] Po ss ibl e answer : [yG inserted here] Is the pos si ble answer : A ) True B ) False Respond with A or B only
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Uncertainty Quantification for LLM Function-Calling
Multi-sample UQ methods like Semantic Entropy show no clear advantage over single-sample methods for LLM function-calling, but both improve when adapted using abstract syntax tree clustering for multi-sample approaches and semantic token selection for logit-based single-sample scores.