Response-Level Rewards

Follow this topic Ask about this topic

Momentum

2 papers in the last four weeks, against 2 the four weeks before. 0.0% of all new papers.

Jul 13Week of Sep 28

Latest papers 23

All topics

MeanFlowAdvantage: Stable Reward Fine-Tuning for Few-Step Average-Velocity Generators
Sep 29, 2026Haocheng Tang, Tianchi Xie, Xingqiao LinMeanflow Generative Flow Networks
Northeastern University Boston, MA 02115, USA · Tsinghua University Beijing, 100084, PRC · Carnegie Mellon University Pittsburgh, PA 15213, USA
From Soft Targets to Reward Signals: How Assignment and Reward Objectives Interact
Sep 28, 2026Jiangtao Lin, Bangyang Wei, Siyi Liu +2Response-Level Rewards Reward Signal
Tsinghua Shenzhen International Graduate School, Tsinghua University · School of Vehicle and Mobility, Tsinghua University · Tencent Holdings Limited +1
Co-Evolving Actor-Conditioned Critics for Non-Verifiable Generation
Aug 31, 2026Jinyoung Kim, Muhammad Khalifa, Lajanugen Logeswaran +4Critique Critic
University of Michigan · LG AI Research · University of Illinois at Chicago
It's a matter of timescale: non-linear utility in successor features and multi-objective planning and learning
Aug 26, 2026Liam P. H. Mertens, Lucas N. Alegre, Florent Delgrange +3Multi-Objective Reinforcement Learning Successor Representations
AI Lab, Vrije Universiteit Brussel, Belgium · Institute of Informatics - Federal University of Rio Grande do Sul · Federation University Australia
A Mean-Field Framework for Inference-Time Distributional Control of Diffusion Models
Aug 9, 2026Samuel Howard, Nikolas NüskenDiffusion Sampling Inference-Time Steering
Department of Statistics, University of Oxford · Department of Mathematics, King’s College London
AVCap: Reinforcing Audio-Video Joint Caption with Detail-Aware Reward
Aug 7, 2026Mingyang Wu, Kaituo Feng, Bohao Li +3Response-Level Rewards
MMLab, CUHK · Peking University
Post-Training at the Edge of Detectability: A Game-Theoretic Approach to Fine-Tuning
Jul 29, 2026Keegan Harris, Brian W. Lee, Ian Waudby-Smith +3Reinforcement Fine-Tuning Model Fine-Tuning
University of California, Berkeley · Inria & École Normale Supérieure
Decentralised Consensus Learning Networks: SME Rotation Without Centralised Reward
Jul 27, 2026Florin NeaguDecentralized Learning Response-Level Rewards
Reward Valuation in Large Language Models: Causal Induction of Anhedonia
Jul 7, 2026Melika Honarmand, Samin Mahdipour Aghabagher, Martin SchrimpfResponse-Level Rewards Intrinsic Motivation
NeuroAI Laboratory, EPFL
Many Voices, One Reward: Multi-Role Rubric Generation for LLM Judging and Reward Modeling
Jul 2, 2026Dazhi Fu, Jiuding Yang, Yiwen Guo +1Task-Specific Rubrics Progress Reward Modeling
School of Data Science, The Chinese University of Hong Kong, Shenzhen, China · LIGHTSPEED · Independent Researcher
UniCoder: Unified Visual-to-Code Generation via Symbolic Rewards and Reference-Guided Code Optimization
Jun 30, 2026Yaozhi Zheng, Yilei Jiang, Manyuan Zhang +5Image-To-Code Generation Visual Generation
The Chinese University of Hong Kong · Shanghai Artificial Intelligence Laboratory
The Verification Horizon: No Silver Bullet for Coding Agent Rewards
Jun 24, 2026Binghai Wang, Chenlong Zhang, Dayiheng Liu +10Coding Agents Response-Level Rewards
Qwen Team
Reinforcement Learning for Computer-Use Agents with Autonomous Evaluation
Jun 23, 2026Marta Sumyk, Oleksandr KosovanGraphical User Interface Agents Computer-Use Agents
Ukrainian Catholic University, Lviv, Ukraine
GIFT: LLM-Guided State-Reward Interface for Financial Reinforcement Learning
Jun 7, 2026Yanyan Wu, Boyi Zhang, Yanlin Liu +10Large Language Model Reinforcement Learning Offline Reinforcement Learning
East China University of Science and Technology · University of Science and Technology of China · Southwestern University of Finance and Economics +5
SAW: Stage-Aware Dynamic Weighting for Multi-Objective Reinforcement Learning in Large Language Models
Jun 5, 2026Yuchen He, Baolong Bi, Shenghua Liu +7Multi-Objective Reinforcement Learning Large Language Model Reinforcement Learning
Institute of Computing Technology, Chinese Academy of Sciences · University of Electronic Science and Technology of China
Verifiable Rewards Beyond Math and Code: Lightweight Corpus-Grounded Process Supervision for Factual Question Answering
May 28, 2026Shicheng Fan, Haochang Hao, Dehai Min +3Verifiable Rewards Text Corpora
University of Illinois Chicago
Plan Before Search: Search Agents Need Plan
May 27, 2026Zhipeng Qian, Zihan Liang, Yufei Ma +7Large Language Model Planning Search Agents
1Kuaishou Technology · 2Key Laboratory of Multimedia Trusted Perception and Efficient Computing, Ministry of Education of China, Xiamen University, 361005, P.R. China · 3VCIP, CS, Nankai University
DelTA: Discriminative Token Credit Assignment for Reinforcement Learning from Verifiable Rewards
May 20, 2026Kaiyi Zhang, Wei Wu, Yankai LinReinforcement Learning With Verifiable Reward Credit Assignment
Gaoling School of Artificial Intelligence, Renmin University of China · Ant International
X-SYNTH: Beyond Retrieval -- Enterprise Context Synthesis from Observed Digital Human Attention
May 15, 2026Guruprasad Raghavan, George Nychis, Rohan Narayana MurthyContext-Awareness Synthesizer
Workfabric AI
Object Hallucination-Free Reinforcement Unlearning for Vision-Language Models
May 8, 2026Kaidi Jia, Yujie Lin, Chengyi Yang +2Recent Vision-Language Models Large Language Model Hallucination
1Xiamen University
Incentivizing Parametric Knowledge via Reinforcement Learning with Verifiable Rewards for Cross-Cultural Entity Translation
Apr 18, 2026Jiang Zhou, Xiaohu Zhao, Xinwei Wu +8Reinforcement Learning With Verifiable Reward Pretraining
TJUNLP Lab, Tianjin University, China · Alibaba Group, China
Region-Constrained Group Relative Policy Optimization for Flow-Based Image Editing
Apr 10, 2026Zhuohan Ouyang, Zhe Qian, Wenhuo Cui +1Image Editing Repair-Based Group-Relative Policy Optimization
1South China Normal University · 2South China Agricultural University · 3Monash University
Explicit Credit Assignment through Local Rewards and Dependence Graphs in Multi-Agent Reinforcement Learning
Jan 29, 2026Bang Giang Le, Viet Cuong TaMulti-Agent Reinforcement Learning Credit Assignment
Human-Machine Interaction Laboratory VNU University of Engineering and Technology, Hanoi, Vietnam

Sep 29, 2026, cs.AIJ/K move · Enter open · S save

MeanFlowAdvantage: Stable Reward Fine-Tuning for Few-Step Average-Velocity Generators

Haocheng Tang, Tianchi Xie, Xingqiao Lin

Northeastern University Boston, MA 02115, USA · Tsinghua University Beijing, 100084, PRC · Carnegie Mellon University Pittsburgh, PA 15213, USA

Meanflow Generative Flow Networks Few-Step Generation Reinforcement Fine-Tuning Response-Level Rewards Model Fine-Tuning

Open paper AskPDF arXiv

MeanFlow enables efficient few-step generation by predicting interval-average velocities, but this representation creates a mismatch for reward fine-tuning: existing advantage-based objectives are typically defined on instantaneous velocities or equivalent

x_0

-space predictions, whereas inference directly uses the learned average-velocity map. We introduce MeanFlowAdvantage, a signed advantage-weighted least-squares objective for average-velocity generators. Our key construction uses a shared, detached MeanFlow derivative correction to express the reward objective in prediction space while making rollout and reference regularization exact penalties on the average-velocity network deployed at inference. The resulting formulation preserves MeanFlow's native few-step sampler and provides a direct mechanism for transferring reward improvements to the deployed flow map. On SD3.5-Medium, MeanFlowAdvantage improves all eight reported metrics over the matched four-step MeanFlowNFT baseline and, with only four NFEs, matches or exceeds the 40-step DiffusionNFT baseline on six of eight metrics. The same objective also transfers to DNA promoter design, where it supports both teacher-free on-policy RL for a generator defined on a manifold and teacher-guided reward-graded distillation, with the latter yielding the lowest one-step Sei profile MSE among the compared configurations.

Momentum

Related topics

Latest papers 23