CoT training in LLM agents improves prompt-action quality more than the advantage of generated reasoning, and selectively masking action supervision improves out-of-domain generalization.
On the Brittle Foundations of ReAct Prompting for Agentic Large Language Models
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.AI 2years
2026 2verdicts
UNVERDICTED 2representative citing papers
Novelty estimation via LLM prompts enables pruning in Tree-of-Thought search, reducing overall token usage on language planning benchmarks.
citing papers explorer
-
Where Do CoT Training Gains Land in LLM based Agents?
CoT training in LLM agents improves prompt-action quality more than the advantage of generated reasoning, and selectively masking action supervision improves out-of-domain generalization.
-
Novelty-based Tree-of-Thought Search for LLM Reasoning and Planning
Novelty estimation via LLM prompts enables pruning in Tree-of-Thought search, reducing overall token usage on language planning benchmarks.