Image de couverture de livre

Mending the Holes: Mitigating Reward Hacking in Reinforcement Learning for Multilingual Translation

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Liu, Yifeng, Ouyang, Siqi, Revanasiddappa, Yatish Hosmane, Li, Lei
Format:	Preprint
Publié:	2026
Sujets:	Computation and Language
Accès en ligne:	Acceder al recurso
Tags:	Ajouter un tag Pas de tags, Soyez le premier à ajouter un tag!

Documents similaires

Directional Alignment Mitigates Reward Hacking in Reinforcement Learning for Language Models
par: Deng, Wenlong, et autres
Publié: (2026)

RRM: Robust Reward Model Training Mitigates Reward Hacking
par: Liu, Tianqi, et autres
Publié: (2024)

Reward Shaping to Mitigate Reward Hacking in RLHF
par: Fu, Jiayi, et autres
Publié: (2025)

CMU's IWSLT 2025 Simultaneous Speech Translation System
par: Ouyang, Siqi, et autres
Publié: (2025)

RASST: Fast Cross-modal Retrieval-Augmented Simultaneous Speech Translation
par: Luo, Jiaxuan, et autres
Publié: (2026)

ODIN: Disentangled Reward Mitigates Hacking in RLHF
par: Chen, Lichang, et autres
Publié: (2024)

Truthful or Fabricated? Using Causal Attribution to Mitigate Reward Hacking in Explanations
par: Ferreira, Pedro, et autres
Publié: (2025)

InfiniSST: Simultaneous Translation of Unbounded Speech with Large Language Model
par: Ouyang, Siqi, et autres
Publié: (2025)

Translation Canvas: An Explainable Interface to Pinpoint and Analyze Translation Systems
par: Dandekar, Chinmay, et autres
Publié: (2024)

When Reward Hacking Rebounds: Understanding and Mitigating It with Representation-Level Signals
par: Wu, Rui, et autres
Publié: (2026)

Mitigating Reward Hacking in RLHF via Advantage Sign Robustness
par: Ono, Shinnosuke, et autres
Publié: (2026)

Temper and Tilt Lead to SLOP: Reward Hacking Mitigation with Inference-Time Alignment
par: Wang, Ye, et autres
Publié: (2026)

Gradient Regularization Prevents Reward Hacking in Reinforcement Learning from Human Feedback and Verifiable Rewards
par: Ackermann, Johannes, et autres
Publié: (2026)

FASST: Fast LLM-based Simultaneous Speech Translation
par: Ouyang, Siqi, et autres
Publié: (2024)

CA*: Addressing Evaluation Pitfalls in Computation-Aware Latency for Simultaneous Speech Translation
par: Xu, Xi, et autres
Publié: (2024)

Specification Self-Correction: Mitigating In-Context Reward Hacking Through Test-Time Refinement
par: Gallego, Víctor
Publié: (2025)

Paths Not Taken: Understanding and Mending the Multilingual Factual Recall Pipeline
par: Lu, Meng, et autres
Publié: (2025)

Detecting and Suppressing Reward Hacking with Gradient Fingerprints
par: Wang, Songtao, et autres
Publié: (2026)

Spontaneous Reward Hacking in Iterative Self-Refinement
par: Pan, Jane, et autres
Publié: (2024)

Mitigating Language Barriers in Education: Developing Multilingual Digital Learning Materials with Machine Translation
par: Poláková, Lucie, et autres
Publié: (2025)

Hierarchical Policy Optimization for Simultaneous Translation of Unbounded Speech
par: Ouyang, Siqi, et autres
Publié: (2026)

Teaching Models to Verbalize Reward Hacking in Chain-of-Thought Reasoning
par: Turpin, Miles, et autres
Publié: (2025)

Anticipating Future with Large Language Model for Simultaneous Machine Translation
par: Ouyang, Siqi, et autres
Publié: (2024)

MERIT: Multilingual Expert-Reward Informed Tuning for Chinese-Centric Low-Resource Machine Translation
par: Lu, Zhixiang, et autres
Publié: (2026)

Alleviating Attention Hacking in Discriminative Reward Modeling through Interaction Distillation
par: Zang, Jianxiang
Publié: (2025)

Balancing Rewards in Text Summarization: Multi-Objective Reinforcement Learning via HyperVolume Optimization
par: Song, Junjie, et autres
Publié: (2025)

SSR-Zero: Simple Self-Rewarding Reinforcement Learning for Machine Translation
par: Yang, Wenjie, et autres
Publié: (2025)

Agentic Reinforcement Learning with Implicit Step Rewards
par: Liu, Xiaoqian, et autres
Publié: (2025)

Rewarding Creativity: A Human-Aligned Generative Reward Model for Reinforcement Learning in Storytelling
par: Li, Zhaoyan, et autres
Publié: (2026)

Incentivizing Parametric Knowledge via Reinforcement Learning with Verifiable Rewards for Cross-Cultural Entity Translation
par: Zhou, Jiang, et autres
Publié: (2026)

Conditions for Catastrophic Forgetting in Multilingual Translation
par: Liu, Danni, et autres
Publié: (2025)

Monitoring Emergent Reward Hacking During Generation via Internal Activations
par: Wilhelm, Patrick, et autres
Publié: (2026)

Multilingual Machine Translation with Large Language Models: Empirical Results and Analysis
par: Zhu, Wenhao, et autres
Publié: (2023)

Is It Thinking or Cheating? Detecting Implicit Reward Hacking by Measuring Reasoning Effort
par: Wang, Xinpeng, et autres
Publié: (2025)

CMU's IWSLT 2024 Simultaneous Speech Translation System
par: Xu, Xi, et autres
Publié: (2024)

Feedback Loops With Language Models Drive In-Context Reward Hacking
par: Pan, Alexander, et autres
Publié: (2024)

ExTrans: Multilingual Deep Reasoning Translation via Exemplar-Enhanced Reinforcement Learning
par: Wang, Jiaan, et autres
Publié: (2025)

Beyond Sparse Rewards: Enhancing Reinforcement Learning with Language Model Critique in Text Generation
par: Cao, Meng, et autres
Publié: (2024)

Learning When to Translate for Multilingual Reasoning
par: Kang, Deokhyung, et autres
Publié: (2026)

Attention Sinks in Massively Multilingual Neural Machine Translation:Discovery, Analysis, and Mitigation
par: Mutisya, Hillary, et autres
Publié: (2026)