ST-Veto: Spatio-Temporal Token Veto for Diffusion MLLMs via Taylor Prediction and Visual Grounding
Ranking
Overall
72
Content
85
Popularity
41
Observed public metrics from 1 member.
Merged summary
TL;DR - ST-Veto is a training-free decoding method for diffusion multimodal LLMs that replaces temporally unstable or weakly image-grounded tokens. It improves multimodal reasoning accuracy by up to 9% without additional training or generation cost.
- Uses second-order Taylor prediction to identify tokens with unstable confidence across diffusion steps.
- Measures image-attention mass to filter tokens lacking strong visual grounding.
- Swaps vetoed tokens with safer candidates during iterative unmasking.
- Consistently outperforms standard decoding policies and prior VLM reasoning methods across multiple models and benchmarks.
Sources (1)
ST-Veto: Spatio-Temporal Token Veto for Diffusion MLLMs via Taylor Prediction and Visual Grounding
Public signals
Semantic Scholar citations 0 · Semantic Scholar influential citations 0
TL;DR - ST-Veto is a training-free decoding method for diffusion multimodal LLMs that replaces temporally unstable or weakly image-grounded tokens. It improves multimodal reasoning accuracy by up to 9% without additional training or generation cost.
- Uses second-order Taylor prediction to identify tokens with unstable confidence across diffusion steps.
- Measures image-attention mass to filter tokens lacking strong visual grounding.
- Swaps vetoed tokens with safer candidates during iterative unmasking.
- Consistently outperforms standard decoding policies and prior VLM reasoning methods across multiple models and benchmarks.