Research map / Image generation / Fast diffusion sampling
dMLLM-TTS: Self-Verified and Efficient Test-Time Scaling for Diffusion Multi-Modal Large Language Models
Yi Xin, Siqi Luo, Tianxiang Xu, Qi Qin, Haoxing Chen, Kaiwen Zhu, Zhiwei Zhang, Yangfan He, Rongchao Zhang, Jinbin Bai, Shuo Cao, Bin Fu and 4 more
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use hierarchical trajectory search instead of linear search (T=32,N=32) (ablation) | dMLLM-TTS · GenEval overall (MMaDA) | accuracy: 0.64 -> 0.66 | Improved |
| use hierarchical trajectory search instead of linear search (T=32,N=32) (ablation) | dMLLM-TTS · GenEval overall (Lumina-DiMOO) | accuracy: 0.91 -> 0.92 | Improved |
| HTS reaches equal scores with less compute (ablation) | dMLLM-TTS · GenEval | NFE: 5x faster on Lumina-DiMOO, 6x on MMaDA and Muddit | Improved |
| apply test-time scaling (HTS+SVF, T=32,N=32) | Lumina-DiMOO · GenEval overall | accuracy: 0.78 -> 0.92 | Improved |
| apply test-time scaling (HTS+SVF, T=32,N=32) | MMaDA · GenEval overall | accuracy: 0.51 -> 0.66 | Improved |
| apply test-time scaling (HTS+SVF, T=32,N=32) | Muddit · GenEval overall | accuracy: 0.53 -> 0.67 | Improved |
| scale trajectories N from 1 to 32 (ablation) | dMLLM-TTS · GenEval overall (Lumina-DiMOO) | accuracy: +8.8% | Improved |
| scale refinement steps from 8 to 64 (ablation) | dMLLM-TTS · GenEval overall | accuracy | Improved |
| replace self-verified feedback with GPT-4o verifier (ablation) | dMLLM-TTS · GenEval overall (Muddit) | accuracy: 0.67 -> 0.74 | Improved |
| replace self-verified feedback with VILA-Judge verifier (ablation) | dMLLM-TTS · GenEval overall | accuracy: Lumina-DiMOO 0.92 -> 0.90; MMaDA 0.66 -> 0.70; Muddit 0.67 -> 0.70 | Mixed |
Closest papers by meaning
- Efficient Test-Time Scaling via Hierarchical Search and Self-Verification for Discrete Diffusion Language Models (ICML 2026 · similarity 0.91)
- Exploring the Deep Fusion of Large Language Models and Diffusion Transformers for Text-to-Image Synthesis (CVPR 2025 · similarity 0.90)
- On the Scalability of Diffusion-based Text-to-Image Generation (CVPR 2024 · similarity 0.89)
- Scaling Inference Time Compute for Diffusion Models (CVPR 2025 · similarity 0.89)
- DAWN: Dependency-Aware Fast Inference for Diffusion LLMs (NeurIPS 2026 · similarity 0.89)
- Thinking Diffusion: Penalize and Guide Visual-Grounded Reasoning in Diffusion Multimodal Language Models (CVPR 2026 · similarity 0.89)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
