GR-GRPO: Graph-Diffused Credit for Autoregressive Image RL Alignment
- 15 September 2026
- book chapter
- Published by Springer Nature
- p. 185-202
Abstract
No abstract availableKeywords
This publication has 22 references indexed in Scilit:
- Infinity∞: Scaling Bitwise AutoRegressive Modeling for High-Resolution Image Synthesis2025
- RandAR: Decoder-only Autoregressive Visual Generation in Random Orders2025
- Teaching Large Language Models to Regress Accurate Image Quality Scores Using Score Distribution2025
- Janus: Decoupling Visual Encoding for Unified Multimodal Understanding and Generation2025
- TreeRL: LLM Reinforcement Learning with On-Policy Tree Search2025
- Using Human Feedback to Fine-tune Diffusion Models without Any Reward Model2024
- Diffusion Model Alignment Using Direct Preference Optimization2024
- Speculative Decoding: Exploiting Speculative Execution for Accelerating Seq2seq Generation2023
- MaskGIT: Masked Generative Image Transformer2022
- Taming Transformers for High-Resolution Image Synthesis2021