paper-with-me

Papers

Penalizing Length: Uncovering Systematic Bias in Quality Estimation Metrics

2025-10-24 · Yilin Zhang, Wenda Xu, Zhongtao Liu, Tetsuji Nakagawa, Markus Freitag arxiv

Quality Estimation (QE) metrics are vital in machine translation for reference-free evaluation and increasingly serve as selection criteria in data filtering and candidate reranking. However, the prevalence and impact of length bias in QE metrics have been underexplored. Through a systematic study of top-performing learned and LLM-as-a-Judge QE metrics across 10 diverse language pairs, we reveal two critical length biases: First, QE metrics consistently over-predict errors with increasing translation length, even for high-quality, error-free texts. Second, they exhibit a systematic preference for shorter translations when multiple candidates of comparable quality are available for the same source text. These biases risk unfairly penalizing longer, correct translations and can propagate into downstream pipelines that rely on QE signals for data selection or system optimization. We trace the root cause of learned QE metrics to skewed supervision distributions, where longer error-free examples are underrepresented in training data. As a diagnostic intervention, we apply length normalization during training and show that this simple modification effectively decouples error prediction from sequence length, yielding more reliable QE signals across translations of varying length.

📄 PDF Abstract BibTeX arXiv:2510.22028

Code (0)

등록된 구현이 없습니다.

Tasks

Machine Translation

Similar Papers 제목 키워드 기반

Beyond Penalizing Mistakes: Stabilizing Efficiency Training in Large Reasoning Models via Adaptive Correct-Only Rewards

2026-06-21 · Jungseob Lee, Seungyoon Lee, Seongtae Hong, Minhyuk Kim 외 arxiv

Training large language models to reason efficiently is a critical challenge. While integrating length-penalizing rewards into Group Relative Policy Optimization (GRPO) aims to reduce verbosity, it frequently triggers re…

Mathematical Reasoning

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

Uncovering Gender Bias in Media Coverage of Politicians with Machine Learning

2020-05-15 · Susan Leavy

This paper presents research uncovering systematic gender bias in the representation of political leaders in the media, using artificial intelligence. Newspaper coverage of Irish ministers over a fifteen year period was …

BIG-bench Machine Learning

EB-GLS: An Improved Guided Local Search Based on the Big Valley Structure

2017-09-22 · Jialong Shi, Qingfu Zhang, Edward Tsang

Local search is a basic building block in memetic algorithms. Guided Local Search (GLS) can improve the efficiency of local search. By changing the guide function, GLS guides a local search to escape from locally optimal…

Combinatorial OptimizationTraveling Salesman Problem

Uncovering Bias Mechanisms in Observational Studies

2025-06-01 · Ilker Demirel, Zeshan Hussain, Piersilvio De Bartolomeis, David Sontag

Observational studies are a key resource for causal inference but are often affected by systematic biases. Prior work has focused mainly on detecting these biases, via sensitivity analyses and comparisons with randomized…

Causal Inference