Research map / Image generation / Text-to-image diffusion
DiverseVAR: Balancing Diversity and Quality of Next-Scale Visual Autoregressive Models
Mingue Park, Prin Phunyaphibarn, Phillip (Yuseung) Lee, Minhyuk Sung
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 |
|---|---|---|---|
| add condition annealing (text-embedding noise) (ablation) | Infinity · MJHQ-30K | LPIPS-MPD: 0.33 -> 0.56 | Improved |
| add condition annealing (text-embedding noise) (ablation) | Infinity · MJHQ-30K | ImageReward: 1.22 -> 0.49 | Hurt |
| add scale-travel refinement after condition annealing (ablation) | Infinity · MJHQ-30K | ImageReward: 0.49 -> 1.08 | Improved |
| add scale-travel refinement after condition annealing (ablation) | Infinity · MJHQ-30K | LPIPS-MPD: 0.56 -> 0.48 | Hurt |
| add scale-travel refinement after condition annealing (ablation) | Infinity · MS-COCO | FID: 22.51 -> 28.97 | Hurt |
| add condition annealing (text-embedding noise) (ablation) | Switti · MJHQ-30K | FID: 16.18 -> 21.34 | Hurt |
| add scale-travel refinement after condition annealing (ablation) | Switti · MJHQ-30K | FID: 21.34 -> 17.30 | Improved |
| add condition annealing (text-embedding noise) (ablation) | InfinityStar · VBench | diversity: 0.270 -> 0.627 | Improved |
| add scale-travel refinement after condition annealing (ablation) | InfinityStar · VBench | quality score: 0.794 -> 0.804 | Improved |
| add scale-travel refinement after condition annealing (ablation) | InfinityStar · VBench | diversity: 0.627 -> 0.487 | Hurt |
| raise sampling temperature from 1 to 10 (ablation) | Infinity · MJHQ-30K | ImageReward: 1.323 -> -0.430 | Hurt |
| raise nucleus sampling p from 0.9 to 1.0 (ablation) | Infinity · MJHQ-30K | LPIPS-MPD: 0.323 -> 0.335 | No clear effect |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- Infinity — Infinity: Scaling Bitwise AutoRegressive Modeling for High-Resolution Image Synthesis (CVPR 2025) · better on 4 of 6 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- FasterVAR: Plug-and-Play Acceleration for Visual Autoregressive Models (ICML 2026 · similarity 0.91)
- Visual Autoregressive Modeling: Scalable Image Generation via Next-Scale Prediction (NeurIPS 2024 · similarity 0.90)
- ARVAR: Accelerating Visual Autoregressive Model via Attention Retrospect (ECCV 2026 · similarity 0.90)
- TTS-VAR: A Test-Time Scaling Framework for Visual Auto-Regressive Generation (NeurIPS 2025 · similarity 0.90)
- FlexVAR: Flexible Visual Autoregressive Modeling without Residual Prediction (NeurIPS 2025 · similarity 0.89)
- Fundamental Limits of Visual Autoregressive Transformers: Universal Approximation Abilities (ICML 2025 · similarity 0.88)
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.
