Image de couverture de livre

Reason Only When Needed: Efficient Generative Reward Modeling via Model-Internal Uncertainty

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Xue, Chao, Wang, Yao, Liu, Mengqiao, Liang, Di, Han, Xingsheng, Liu, Peiyang, Wu, Xianjie, Lu, Chenyao, Jiang, Lei, Lu, Yu, Shi, Haibo, Liang, Shuang, Peng, Minlong, Salim, Flora D.
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

Why Supervised Fine-Tuning Fails to Learn: A Systematic Study of Incomplete Learning in Large Language Models
par: Xue, Chao, et autres
Publié: (2026)

Parameter Importance is Not Static: Evolving Parameter Isolation for Supervised Fine-Tuning
par: Lin, Zekai, et autres
Publié: (2026)

Structural Reward Model: Enhancing Interpretability, Efficiency, and Scalability in Reward Modeling
par: Liu, Xiaoyu, et autres
Publié: (2025)

DeCoRL: Decoupling Reasoning Chains via Parallel Sub-Step Generation and Cascaded Reinforcement for Interpretable and Scalable RLHF
par: Gao, Ziyuan, et autres
Publié: (2025)

Not All Parameters Are Created Equal: Smart Isolation Boosts Fine-Tuning Performance
par: Wang, Yao, et autres
Publié: (2025)

Think Only When You Need with Large Hybrid-Reasoning Models
par: Jiang, Lingjie, et autres
Publié: (2025)

Mechanistic Indicators of Steering Effectiveness in Large Language Models
par: Jafari, Mehdi, et autres
Publié: (2026)

When You Need to Know
par: Brown, Flora Morris
Publié: (1976)

Think-RM: Enabling Long-Horizon Reasoning in Generative Reward Models
par: Hong, Ilgee, et autres
Publié: (2025)

Self-Rewarding PPO: Aligning Large Language Models with Demonstrations Only
par: Zhang, Qingru, et autres
Publié: (2025)

Reinforcement Learning Enhanced Multi-hop Reasoning for Temporal Knowledge Question Answering
par: Wen, Wuzhenghong, et autres
Publié: (2026)

Whether Uncertainty Theory Can Enhance GDP Forecasting From Energy: A New Uncertain MIDAS Model
par: Yuxin Shi, et autres
Publié: (2025)

DeepLévy: Learning Heavy-Tailed Uncertainty in Highly Volatile Time Series
par: Yang, Yang, et autres
Publié: (2026)

Uncertainty Quantification for Large Language Model Reward Learning under Heterogeneous Human Feedback
par: Liu, Pangpang, et autres
Publié: (2025)

DPI: Exploiting Parameter Heterogeneity for Interference-Free Fine-Tuning
par: Liu, Xiaoyu, et autres
Publié: (2026)

Reward Model Routing in Alignment
par: Wu, Xinle, et autres
Publié: (2025)

MAPLE: Mobile App Prediction Leveraging Large Language Model Embeddings
par: Khaokaew, Yonchanok, et autres
Publié: (2023)

Who Stole Your Data? A Method for Detecting Unauthorized RAG Theft
par: Liu, Peiyang, et autres
Publié: (2025)

STORM: Internalized Modeling for Spatial-Temporal Reasoning in Video-Language Models
par: Liang, Yiming, et autres
Publié: (2026)

Ask Only When Needed: Proactive Retrieval from Memory and Skills for Experience-Driven Lifelong Agents
par: Cai, Yuxuan, et autres
Publié: (2026)

Take Only What You Need: Rank Minimization as an Implicit Forgetting Regularizer in Continual Learning
par: Lu, Haodong, et autres
Publié: (2024)

Large Language Models are Contrastive Reasoners
par: Yao, Liang
Publié: (2024)

Optimal Uncertainty Quantification under General Moment Constraints on Input Subdomains
par: Jin, Rong, et autres
Publié: (2025)

Double-Diffusion: ODE-Prior Accelerated Diffusion Models for Spatio-Temporal Graph Forecasting
par: Dong, Hanlin, et autres
Publié: (2025)

A-UTE: Advection Informed, Uncertainty Aware Temperature Emulator
par: Saleem, Hira, et autres
Publié: (2024)

Large Language Models for Next Point-of-Interest Recommendation
par: Li, Peibo, et autres
Publié: (2024)

When Ads Become Profiles: Uncovering the Invisible Risk of Web Advertising at Scale with LLMs
par: Chen, Baiyu, et autres
Publié: (2025)

Decoder-Only LLMs are Better Controllers for Diffusion Models
par: Dong, Ziyi, et autres
Publié: (2025)

STAR: Failure-Aware Markovian Routing for Multi-Agent Spatiotemporal Reasoning
par: Yang, Ruiyi, et autres
Publié: (2026)

From XXLTraffic to EvoXXLTraffic: Scaling Traffic Forecasting to Sensor-Evolving Networks
par: Yin, Du, et autres
Publié: (2026)

XXLTraffic: Expanding and Extremely Long Traffic forecasting beyond test adaptation
par: Yin, Du, et autres
Publié: (2024)

HiT-JEPA: A Hierarchical Self-supervised Trajectory Embedding Framework for Similarity Computation
par: Li, Lihuan, et autres
Publié: (2025)

Large Language Model Reasoning Failures
par: Song, Peiyang, et autres
Publié: (2026)

Learning from Contrasts: Synthesizing Reasoning Paths from Diverse Search Trajectories
par: Liu, Peiyang, et autres
Publié: (2026)

Free(): Learning to Forget in Malloc-Only Reasoning Models
par: Zheng, Yilun, et autres
Publié: (2026)

EMAC+: Embodied Multimodal Agent for Collaborative Planning with VLM+LLM
par: Ao, Shuang, et autres
Publié: (2025)

ZARA: Training-Free Motion Time-Series Reasoning via Evidence-Grounded LLM Agents
par: Li, Zechen, et autres
Publié: (2025)

Native Reasoning Models: Training Language Models to Reason on Unverifiable Data
par: Wang, Yuanfu, et autres
Publié: (2026)

Joint Reward Modeling: Internalizing Chain-of-Thought for Efficient Visual Reward Models
par: Yang, Yankai, et autres
Publié: (2026)

When Safety Becomes a Vulnerability: Exploiting LLM Alignment Homogeneity for Transferable Blocking in RAG
par: Li, Junchen, et autres
Publié: (2026)