Cover Image

Rethinking the Comparison Unit in Sequence-Level Reinforcement Learning: An Equal-Length Paired Training Framework from Loss Correction to Sample Construction

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Ding, Fei, Zhang, Yongkang, Liu, Runhao, Liao, Yuhao, Zeng, Zijian, Yang, Huiming, wang, Sibo, Liao, Linglin
Format:	Preprint
Published:	2026
Subjects:	Machine Learning Artificial Intelligence
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

Internalizing Outcome Supervision into Process Supervision: A New Paradigm for Reinforcement Learning for Reasoning
by: Ding, Fei, et al.
Published: (2026)

PoseStreamer: A Multi-modal Framework for 3D Tracking of Unseen Moving Objects
by: Yang, Huiming, et al.
Published: (2025)

Difference Feedback: Generating Multimodal Process-Level Supervision for VLM Reinforcement Learning
by: Feiding, et al.
Published: (2026)

Design Conditions for Intra-Group Learning of Sequence-Level Rewards: Token Gradient Cancellation
by: Ding, Fei, et al.
Published: (2026)

DexSim2Real: Foundation Model-Guided Sim-to-Real Transfer for Generalizable Dexterous Manipulation
by: Zeng, Zijian, et al.
Published: (2026)

Spatial-aware Symmetric Alignment for Text-guided Medical Image Segmentation
by: Liao, Linglin, et al.
Published: (2025)

HELM: Harness-Enhanced Long-horizon Memory for Vision-Language-Action Manipulation
by: Zeng, Zijian, et al.
Published: (2026)

Reducing Credit Assignment Variance via Counterfactual Reasoning Paths
by: Ding, Fei, et al.
Published: (2026)

Not All LLM-Generated Data Are Equal: Rethinking Data Weighting in Text Classification
by: Kuo, Hsun-Yu, et al.
Published: (2024)

Hyperon Pair Production at BESIII
by: Shang, Zijie, et al.
Published: (2025)

Balancing the Reasoning Load: Difficulty-Differentiated Policy Optimization with Length Redistribution for Efficient and Robust Reinforcement Learning
by: Xia, Yinan, et al.
Published: (2026)

Sequence-Level Unsupervised Training in Speech Recognition: A Theoretical Study
by: Yang, Zijian, et al.
Published: (2026)

Rethinking External Slow-Thinking: From Snowball Errors to Probability of Correct Reasoning
by: Gan, Zeyu, et al.
Published: (2025)

ShareDP: Finding k Disjoint Paths for Multiple Vertex Pairs
by: Yuan, Zhiqiu, et al.
Published: (2025)

Memory Sequence Length of Data Sampling Impacts the Adaptation of Meta-Reinforcement Learning Agents
by: Zhang, Menglong, et al.
Published: (2024)

The equivalent condition for GRL codes to be MDS, AMDS or self-dual
by: Liang, Zhonghao, et al.
Published: (2025)

The asymptotic estimation for two classes of generalized Fibonacci sub-sequences
by: Wan, Yongkang, et al.
Published: (2025)

The inverse of the (alternating) infinite sum of the reciprocal of the weighted sum for generalized Fibonacci sub-sequences
by: Wan, Yongkang, et al.
Published: (2025)

Not All Negative Samples Are Equal: LLMs Learn Better from Plausible Reasoning
by: Di, Zixiang, et al.
Published: (2026)

On Duplication-Free Codes for Disjoint or Equal-Length Errors
by: Yu, Wenjun, et al.
Published: (2024)

Clip Your Sequences Fairly: Enforcing Length Fairness for Sequence-Level RL
by: Mao, Hanyi, et al.
Published: (2025)

Multi-Layer GRPO: Enhancing Reasoning and Self-Correction in Large Language Models
by: Ding, Fei, et al.
Published: (2025)

Constructing $k$-ary Orientable Sequences with Asymptotically Optimal Length
by: Gabrić, Daniel, et al.
Published: (2024)

Broad Band Equal-Length And Equal-Width Substrate Integrated Waveguide Four Channel Power Divider
by: Masoud Khoubroo Eslamloo
Published: (2015)

Correction to ‘Rethinking Dopamine‐Guided Action Sequence Learning’
Published: (2025)

sprofing the DLSCA
by: wang
Published: (2025)

Touch to Pair: Secure and Usable IoT Pairing without Information Loss
by: Wu, Chuxiong, et al.
Published: (2024)

Adversarial Samples Are Not Created Equal
by: Crawford, Jennifer, et al.
Published: (2026)

AI for Equitable Tennis Training: Leveraging AI for Equitable and Accurate Classification of Tennis Skill Levels and Training Phases
by: Gao, Gyanna, et al.
Published: (2024)

Conversation Disentanglement with Bi-Level Contrastive Learning
by: Huang, Chengyu, et al.
Published: (2022)

Box-Level Class-Balanced Sampling for Active Object Detection
by: Liao, Jingyi, et al.
Published: (2025)

From Production Envelopes to Executable Schedules: Sound Constructive Refinement for High-Mix Manufacturing
by: Liu, Runhao, et al.
Published: (2025)

GDEPO: Group Dual-dynamic and Equal-right Advantage Policy Optimization with Enhanced Training Data Utilization for Sample-Constrained Reinforcement Learning
by: Yan, Zhengqing, et al.
Published: (2026)

Artificial Intelligence in Power System Security and Stability Analysis: A Comprehensive Review
by: Zhang, Runhao
Published: (2024)

Selective Matching Losses -- Not All Scores Are Created Equal
by: Shamir, Gil I., et al.
Published: (2025)

A Fast Algorithm for Scheduling Equal-Length Jobs on Identical Machines
by: Nodari Vakhania
Published: (1998)

Length-Unbiased Sequence Policy Optimization: Revealing and Controlling Response Length Variation in RLVR
by: Liu, Fanfan, et al.
Published: (2026)

Rethinking Model Evaluation as Narrowing the Socio-Technical Gap
by: Liao, Q. Vera, et al.
Published: (2023)

Power-SMC: Low-Latency Sequence-Level Power Sampling for Training-Free LLM Reasoning
by: Azizi, Seyedarmin, et al.
Published: (2026)

Dataset Decomposition: Faster LLM Training with Variable Sequence Length Curriculum
by: Pouransari, Hadi, et al.
Published: (2024)