CAST extracts complexity and failure profiles from historical tool-use trajectories to drive adaptive reasoning and fine-grained rewards in RL, yielding up to 5.85 pp higher execution accuracy and 26% shorter reasoning on BFCLv2 and ToolBench.
In: Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Case-Based Calibration of Adaptive Reasoning and Execution for LLM Tool Use
CAST extracts complexity and failure profiles from historical tool-use trajectories to drive adaptive reasoning and fine-grained rewards in RL, yielding up to 5.85 pp higher execution accuracy and 26% shorter reasoning on BFCLv2 and ToolBench.