Imagen de Portada

When Importance Sampling Misallocates Credit: Asymmetric Ratios for Outcome-Supervised RL

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Wang, Jiakang, Liu, Runze, Cai, Qingpeng, Lin, Lei, Hu, Wenping, Li, Xiu, Zhang, Fuzheng, Zhou, Guorui, Gai, Kun, Pan, Ling
Formato:	Preprint
Publicado:	2025
Materias:	Computation and Language
Acceso en línea:	Acceder al recurso
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

Ejemplares similares

Attention as a Compass: Efficient Exploration for Process-Supervised RL in Reasoning Models
por: Liu, Runze, et al.
Publicado: (2025)

Stabilizing Knowledge, Promoting Reasoning: Dual-Token Constraints for RLVR
por: Wang, Jiakang, et al.
Publicado: (2025)

CE-GPPO: Coordinating Entropy via Gradient-Preserving Clipping Policy Optimization in Reinforcement Learning
por: Su, Zhenpeng, et al.
Publicado: (2025)

Entropy Ratio Clipping as a Soft Global Constraint for Stable Reinforcement Learning
por: Su, Zhenpeng, et al.
Publicado: (2025)

Klear-Reasoner: Advancing Reasoning Capability via Gradient-Preserving Clipping Policy Optimization
por: Su, Zhenpeng, et al.
Publicado: (2025)

Misallocation in the Chinese land market
por: Xuan Fei, et al.
Publicado: (2024)

AURO: Reinforcement Learning for Adaptive User Retention Optimization in Recommender Systems
por: Xue, Zhenghai, et al.
Publicado: (2023)

Leanabell-Prover: Posttraining Scaling in Formal Reasoning
por: Zhang, Jingyuan, et al.
Publicado: (2025)

Random Policy Evaluation Uncovers Policies of Generative Flow Networks
por: He, Haoran, et al.
Publicado: (2024)

Hierarchical Semantic RL: Tackling the Problem of Dynamic Action Space for RL-based Recommendations
por: Wang, Minmao, et al.
Publicado: (2025)

State Regularized Policy Optimization on Data with Dynamics Shift
por: Xue, Zhenghai, et al.
Publicado: (2023)

AIS: Adaptive Importance Sampling for Quantized RL
por: Zhou, Jiajun, et al.
Publicado: (2026)

DISA: Offline Importance Sampling for Distribution-Matching LLM-RL
por: Wang, Shaobo, et al.
Publicado: (2026)

Leanabell-Prover-V2: Verifier-integrated Reasoning for Formal Theorem Proving via Reinforcement Learning
por: Ji, Xingguang, et al.
Publicado: (2025)

Chaos and Misallocation under Price Controls
por: Albrecht, Brian C., et al.
Publicado: (2026)

Symposium on Misallocation and Structural Transformation: Introduction
por: Tasso Adamopoulos, et al.
Publicado: (2024)

Production Function Estimation With Resource Misallocation
por: Shigang Li, et al.
Publicado: (2026)

Inductive-Deductive Strategy Reuse for Multi-Turn Instructional Dialogues
por: Ou, Jiao, et al.
Publicado: (2024)

ERABAL: Enhancing Role-Playing Agents through Boundary-Aware Learning
por: Tang, Yihong, et al.
Publicado: (2024)

Enhancing Role-playing Systems through Aggressive Queries: Evaluation and Improvement
por: Tang, Yihong, et al.
Publicado: (2024)

HoME: Hierarchy of Multi-Gate Experts for Multi-Task Learning at Kuaishou
por: Wang, Xu, et al.
Publicado: (2024)

The Impact of New Digital Infrastructure on Resource Misallocation
por: Qunli Wang, et al.
Publicado: (2026)

MISS: Multi-Modal Tree Indexing and Searching with Lifelong Sequential Behavior for Retrieval Recommendation
por: Guo, Chengcheng, et al.
Publicado: (2025)

The Cancellation Hypothesis in Critic-Free RL: From Outcome Rewards to Token Credits
por: Cheng, Tianhao, et al.
Publicado: (2026)

Future Impact Decomposition in Request-level Recommendations
por: Wang, Xiaobei, et al.
Publicado: (2024)

Sequential Recommendation for Optimizing Both Immediate Feedback and Long-term Retention
por: Liu, Ziru, et al.
Publicado: (2024)

Video Object Segmentation with Dynamic Query Modulation
por: Zhou, Hantao, et al.
Publicado: (2024)

Tournament-Based Performance Evaluation and Systematic Misallocation: Why Forced Ranking Systems Produce Random Outcomes
por: McEntire, Jeremy
Publicado: (2025)

Bifurcated Generative Flow Networks
por: Li, Chunhui, et al.
Publicado: (2024)

How Metro Expansion Influences Enterprise Labor Misallocation
por: Mengting Zhang, et al.
Publicado: (2026)

Random Policy Valuation is Enough for LLM Reasoning with Verifiable Rewards
por: He, Haoran, et al.
Publicado: (2025)

FIM: Frequency-Aware Multi-View Interest Modeling for Local-Life Service Recommendation
por: Wang, Guoquan, et al.
Publicado: (2025)

DialogBench: Evaluating LLMs as Human-like Dialogue Systems
por: Ou, Jiao, et al.
Publicado: (2023)

Just Ask One More Time! Self-Agreement Improves Reasoning of Language Models in (Almost) All Scenarios
por: Lin, Lei, et al.
Publicado: (2023)

Off-policy Distributional Q($λ$): Distributional RL without Importance Sampling
por: Tang, Yunhao, et al.
Publicado: (2024)

CRM: Retrieval Model with Controllable Condition
por: Liu, Chi, et al.
Publicado: (2024)

PROMISE: Process Reward Models Unlock Test-Time Scaling Laws in Generative Recommendations
por: Guo, Chengcheng, et al.
Publicado: (2026)

From Principles to Applications: A Comprehensive Survey of Discrete Tokenizers in Generation, Comprehension, Recommendation, and Information Retrieval
por: Jia, Jian, et al.
Publicado: (2025)

The RNA-seq raw data
por: Pan, Wenping
Publicado: (2030)

A Step Back: Prefix Importance Ratio Stabilizes Policy Optimization
por: Lei, Shiye, et al.
Publicado: (2026)