Cover Image

Fixing the Broken Compass: Diagnosing and Improving Inference-Time Reward Modeling

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Li, Jiachun, Cao, Pengfei, Jin, Zhuoran, Chen, Yubo, Xu, Jiexin, Li, Huaijun, Jiang, Xiaojian, Liu, Kang, Zhao, Jun
Format:	Preprint
Published:	2025
Subjects:	Computation and Language Artificial Intelligence
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

Towards Better Chain-of-Thought: A Reflection on Effectiveness and Faithfulness
by: Li, Jiachun, et al.
Published: (2024)

Cutting Off the Head Ends the Conflict: A Mechanism for Interpreting and Mitigating Knowledge Conflicts in Language Models
by: Jin, Zhuoran, et al.
Published: (2024)

LINKED: Eliciting, Filtering and Integrating Knowledge in Large Language Model for Commonsense Reasoning
by: Li, Jiachun, et al.
Published: (2024)

AgentsCourt: Building Judicial Decision-Making Agents with Court Debate Simulation and Legal Knowledge Augmentation
by: He, Zhitao, et al.
Published: (2024)

Tug-of-War Between Knowledge: Exploring and Resolving Knowledge Conflicts in Retrieval-Augmented Language Models
by: Jin, Zhuoran, et al.
Published: (2024)

Omni-Reward: Towards Generalist Omni-Modal Reward Modeling with Free-Form Preferences
by: Jin, Zhuoran, et al.
Published: (2025)

MIRAGE: Evaluating and Explaining Inductive Reasoning Process in Language Models
by: Li, Jiachun, et al.
Published: (2024)

Agent-RewardBench: Towards a Unified Benchmark for Reward Modeling across Perception, Planning, and Safety in Real-World Multimodal Agents
by: Men, Tianyi, et al.
Published: (2025)

RAG-RewardBench: Benchmarking Reward Models in Retrieval Augmented Generation for Preference Alignment
by: Jin, Zhuoran, et al.
Published: (2024)

Focus on Your Question! Interpreting and Mitigating Toxic CoT Problems in Commonsense Reasoning
by: Li, Jiachun, et al.
Published: (2024)

MMR-Life: Piecing Together Real-life Scenes for Multimodal Multi-image Reasoning
by: Li, Jiachun, et al.
Published: (2026)

RWKU: Benchmarking Real-World Knowledge Unlearning for Large Language Models
by: Jin, Zhuoran, et al.
Published: (2024)

Towards Robust Knowledge Unlearning: An Adversarial Framework for Assessing and Improving Unlearning Robustness in Large Language Models
by: Yuan, Hongbang, et al.
Published: (2024)

Unlocking the Future: Exploring Look-Ahead Planning Mechanistic Interpretability in Large Language Models
by: Men, Tianyi, et al.
Published: (2024)

MMR-V: What's Left Unsaid? A Benchmark for Multimodal Deep Reasoning in Videos
by: Zhu, Kejian, et al.
Published: (2025)

One Mind, Many Tongues: A Deep Dive into Language-Agnostic Knowledge Neurons in Large Language Models
by: Cao, Pengfei, et al.
Published: (2024)

Beyond Under-Alignment: Atomic Preference Enhanced Factuality Tuning for Large Language Models
by: Yuan, Hongbang, et al.
Published: (2024)

A Troublemaker with Contagious Jailbreak Makes Chaos in Honest Towns
by: Men, Tianyi, et al.
Published: (2024)

CITI: Enhancing Tool Utilizing Ability in Large Language Models without Sacrificing General Performance
by: Hao, Yupu, et al.
Published: (2024)

Whispers that Shake Foundations: Analyzing and Mitigating False Premise Hallucinations in Large Language Models
by: Yuan, Hongbang, et al.
Published: (2024)

DTELS: Towards Dynamic Granularity of Timeline Summarization
by: Zhang, Chenlong, et al.
Published: (2024)

Evaluating Personalized Tool-Augmented LLMs from the Perspectives of Personalization and Proactivity
by: Hao, Yupu, et al.
Published: (2025)

Zero-Shot Cross-Lingual Document-Level Event Causality Identification with Heterogeneous Graph Contrastive Transfer Learning
by: He, Zhitao, et al.
Published: (2024)

Knowledge in Superposition: Unveiling the Failures of Lifelong Knowledge Editing for Large Language Models
by: Hu, Chenhui, et al.
Published: (2024)

Towards Atoms of Large Language Models
by: Hu, Chenhui, et al.
Published: (2025)

SemEval-2026 Task 12: Abductive Event Reasoning: Towards Real-World Event Causal Inference for Large Language Models
by: Cao, Pengfei, et al.
Published: (2026)

Revealing the Deceptiveness of Knowledge Editing: A Mechanistic Analysis of Superficial Editing
by: Xie, Jiakuan, et al.
Published: (2025)

WilKE: Wise-Layer Knowledge Editor for Lifelong Knowledge Editing
by: Hu, Chenhui, et al.
Published: (2024)

Knowledge Localization: Mission Not Accomplished? Enter Query Localization!
by: Chen, Yuheng, et al.
Published: (2024)

MEMLA: Enhancing Multilingual Knowledge Editing with Neuron-Masked Low-Rank Adaptation
by: Xie, Jiakuan, et al.
Published: (2024)

CompassJudger-2: Towards Generalist Judge Model via Verifiable Rewards
by: Zhang, Taolin, et al.
Published: (2025)

Inference-Time Scaling for Generalist Reward Modeling
by: Liu, Zijun, et al.
Published: (2025)

Think While Watching: Online Streaming Segment-Level Memory for Multi-Turn Video Reasoning in Multimodal Large Language Models
by: Wang, Lu, et al.
Published: (2026)

Cracking Factual Knowledge: A Comprehensive Analysis of Degenerate Knowledge Neurons in Large Language Models
by: Chen, Yuheng, et al.
Published: (2024)

Continual Few-shot Event Detection via Hierarchical Augmentation Networks
by: Zhang, Chenlong, et al.
Published: (2024)

FinRAGBench-V: A Benchmark for Multimodal RAG with Visual Citation in the Financial Domain
by: Zhao, Suifeng, et al.
Published: (2025)

Reward Auditor: Inference on Reward Modeling Suitability in Real-World Perturbed Scenarios
by: Zang, Jianxiang, et al.
Published: (2025)

RULE: Reinforcement UnLEarning Achieves Forget-Retain Pareto Optimality
by: Zhang, Chenlong, et al.
Published: (2025)

CogMG: Collaborative Augmentation Between Large Language Model and Knowledge Graph
by: Zhou, Tong, et al.
Published: (2024)

The Knowledge Microscope: Features as Better Analytical Lenses than Neurons
by: Chen, Yuheng, et al.
Published: (2025)