Tag
该论文将评估器引导的测试时推理建模为一个计算感知的三动作控制器（trust、fallback、expand），通过学习并校准动作特定的置信度估计来优化推理计算分配，在 Qwen3-8B 上实现约 30% 和 16% 的计算节省。
This study evaluates whether test-time reasoning in large language models (like DeepSeek, GPT, and Gemini) improves net portfolio returns in trading, finding that additional reasoning does not reliably enhance economic outcomes across various conditions.
The paper recasts test-time reasoning methods as recursion operators—Grow, Prune, and Branch—and evaluates them, finding that Branch consistently improves accuracy by recovering from budget-exhausted outputs, while recommending paired scoring for comparative studies.
Claim-Level Reliability Assessment (CLR) is a training-free framework that improves reasoning accuracy in LLMs by verifying critical claims instead of sampling more solutions, reducing token use while boosting performance.