paper-with-me

홈 › Papers

Explaining Length Bias in LLM-Based Preference Evaluations

2024-07-01 · Zhengyu Hu, Linxin Song, Jieyu Zhang, Zheyuan Xiao, Tianfu Wang, Zhengyu Chen, Nicholas Jing Yuan, Jianxun Lian, Kaize Ding, Hui Xiong

The use of large language models (LLMs) as judges, particularly in preference comparisons, has become widespread, but this reveals a notable bias towards longer responses, undermining the reliability of such evaluations. To better understand such bias, we propose to decompose the preference evaluation metric, specifically the win rate, into two key components: desirability and information mass, where the former is length-independent and related to trustworthiness such as correctness, toxicity, and consistency, and the latter is length-dependent and represents the amount of information in the response. We empirically demonstrated the decomposition through controlled experiments and found that response length impacts evaluations by influencing information mass. To derive a reliable evaluation metric that assesses content quality without being confounded by response length, we propose AdapAlpaca, a simple yet effective adjustment to win rate measurement. Specifically, AdapAlpaca ensures a fair comparison of response quality by aligning the lengths of reference and test model responses under equivalent length intervals.

📄 PDF Abstract BibTeX arXiv:2407.01085

Code (0)

등록된 구현이 없습니다.

Tasks

Language ModellingLarge Language Model

Methods 이 논문이 사용한 방법론

DPO 설명 없음

Similar Papers 제목 키워드 기반

Length-Controlled AlpacaEval: A Simple Way to Debias Automatic Evaluators

2024-04-06 · Yann Dubois, Balázs Galambosi, Percy Liang, Tatsunori B. Hashimoto

LLM-based auto-annotators have become a key component of the LLM development process due to their cost-effectiveness and scalability compared to human-based evaluation. However, these auto-annotators can introduce comple…

Chatbotcounterfactual

Post-hoc Reward Calibration: A Case Study on Length Bias

2024-09-25 · Zeyu Huang, Zihan Qiu, Zili Wang, Edoardo M. Ponti 외

Reinforcement Learning from Human Feedback aligns the outputs of Large Language Models with human values and preferences. Central to this process is the reward model (RM), which translates human feedback into training si…

Flattery, Fluff, and Fog: Diagnosing and Mitigating Idiosyncratic Biases in Preference Models

2025-06-05 · Anirudh Bharadwaj, Chaitanya Malaviya, Nitish Joshi, Mark Yatskar

Language models serve as proxies for human preference judgements in alignment and evaluation, yet they exhibit systematic miscalibration, prioritizing superficial patterns over substantive qualities. This bias manifests …

counterfactualData Augmentation

Eliminating Biased Length Reliance of Direct Preference Optimization via Down-Sampled KL Divergence

2024-06-16 · Junru Lu, Jiazheng Li, Siyu An, Meng Zhao 외

Direct Preference Optimization (DPO) has emerged as a prominent algorithm for the direct and robust alignment of Large Language Models (LLMs) with human preferences, offering a more straightforward alternative to the com…

Mitigating Length Bias in RLHF through a Causal Lens

2025-11-16 · Hyeonji Kim, Sujeong Oh, Sanghack Lee arxiv

Reinforcement learning from human feedback (RLHF) is widely used to align large language models (LLMs) with human preferences. However, RLHF-trained reward models often exhibit length bias -- a systematic tendency to fav…

Reinforcement LearningData Augmentation