Recursive Self-Improvement in Unified Multimodal Models
cs.CL, cs.CV
Submitted: 2026-10-02
Updated: 2026-10-02
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Sources
- Chameleon: Mixed-Modal Early-Fusion Foundation Models
- Emerging Properties in Unified Multimodal Pretraining
- EVE: Verifiable Self-Evolution of MLLMs via Executable Visual Transformations
- SUDER: Self-Improving Unified Large Multimodal Models for Understanding and Generation with Dual Self-Rewards
- SRUM: Fine-Grained Self-Rewarding for Unified Multimodal Models
- DataEnvGym: Data Generation Agents in Teacher Environments with Student Feedback
- Zebra-CoT: A Dataset for Interleaved Vision Language Reasoning
- BizGenEval: A Systematic Benchmark for Commercial Visual Content Generation
- UniRL: Self-Improving Unified Multimodal Models via Supervised and Reinforcement Learning
- Ask, Solve, Generate: Self-Evolving Unified Multimodal Understanding and Generation via Self-Consistency Rewards
- Janus: Decoupling Visual Encoding for Unified Multimodal Understanding and Generation
- Show-o: One Single Transformer to Unify Multimodal Understanding and Generation
- DataEvolver: Self-Evolving Multi-Agent Data Construction for Text-Rich Image Generation
- Self-Rewarding Language Models
- STaR: Bootstrapping Reasoning With Reasoning
- EvolveDirector: Approaching Advanced Text-to-Image Generation with Large Vision-Language Models
- Transfusion: Predict the Next Token and Diffuse Images with One Multi-Modal Model
- Factuality Matters: When Image Generation and Editing Meet Structured Visuals
Related papers
- Exploring Solution Divergence and Its Effect on Large Language Model Problem Solving
- Ishigaki-IDS-Bench: A Benchmark for Generating Information Delivery Specification from BIM Information Requirements
- Subliminal Steering: Stronger Encoding of Hidden Signals
- MedStruct-S: A Benchmark for Key Discovery, Key-Conditioned QA and Semi-Structured Extraction from OCR Clinical Reports
- The End of Transformers? On Challenging Attention and the Rise of Sub-Quadratic Architectures
- Untangling the Mechanisms of Misleading Context in Medical Question Answering