paper-with-me

홈 › Papers

Through the Judge's Eyes: Inferred Thinking Traces Improve Reliability of LLM Raters

2025-10-29 · Xingjian Zhang, Tianhong Gao, Suliang Jin, Tianhao Wang, Teng Ye, Eytan Adar, Qiaozhu Mei arxiv

Large language models (LLMs) are increasingly used as raters for evaluation tasks. However, their reliability is often limited for subjective tasks, when human judgments involve subtle reasoning beyond annotation labels. Thinking traces, the reasoning behind a judgment, are highly informative but challenging to collect and curate. We present a human-LLM collaborative framework to infer thinking traces from label-only annotations. The proposed framework uses a simple and effective rejection sampling method to reconstruct these traces at scale. These inferred thinking traces are applied to two complementary tasks: (1) fine-tuning open LLM raters; and (2) synthesizing clearer annotation guidelines for proprietary LLM raters. Across multiple datasets, our methods lead to significantly improved LLM-human agreement. Additionally, the refined annotation guidelines increase agreement among different LLM models. These results suggest that LLMs can serve as practical proxies for otherwise unrevealed human thinking traces, enabling label-only corpora to be extended into thinking-trace-augmented resources that enhance the reliability of LLM raters.

📄 PDF Abstract BibTeX arXiv:2510.25860

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Think-J: Learning to Think for Generative LLM-as-a-Judge

2025-05-20 · Hui Huang, Yancheng He, Hongli Zhou, Rui Zhang 외

LLM-as-a-Judge refers to the automatic modeling of preferences for responses generated by Large Language Models (LLMs), which is of significant importance for both LLM evaluation and reward modeling. Although generative …

Offline RLReinforcement Learning (RL)

Learning to Plan & Reason for Evaluation with Thinking-LLM-as-a-Judge

2025-01-30 · Swarnadeep Saha, Xian Li, Marjan Ghazvininejad, Jason Weston 외

LLM-as-a-Judge models generate chain-of-thought (CoT) sequences intended to capture the step-bystep reasoning process that underlies the final evaluation of a response. However, due to the lack of human annotated CoTs fo…

DeepEyes: Incentivizing "Thinking with Images" via Reinforcement Learning

2025-05-20 · Ziwei Zheng, Michael Yang, Jack Hong, Chenxiao Zhao 외

Large Vision-Language Models (VLMs) have shown strong capabilities in multimodal understanding and reasoning, yet they are primarily constrained by text-based reasoning processes. However, achieving seamless integration …

HallucinationMathematical ReasoningMultimodal Reasoningreinforcement-learning+2

Generating Data-Driven Reasoning Rubrics for Domain-Adaptive Reward Modeling

2026-02-06 · Kate Sanders, Nathaniel Weir, Sapana Chaudhary, Kaj Bostrom 외 arxiv

An impediment to using Large Language Models (LLMs) for reasoning output verification is that LLMs struggle to reliably identify errors in thinking traces, particularly in long outputs, domains requiring expert knowledge…

Reinforcement Learning

InCoder-32B-Thinking: Industrial Code World Model for Thinking

2026-04-03 · Jian Yang, Wei Zhang, Jiajun Wu, Junhang Cheng 외 arxiv

Industrial software development across chip design, GPU optimization, and embedded systems lacks expert reasoning traces showing how engineers reason about hardware constraints and timing semantics. In this work, we prop…