LoopMDM loops early-middle layers in masked diffusion models to match same-size MDM performance with up to 3.3x fewer training FLOPs and outperform on reasoning tasks by up to 8.5 points on GSM8K.
Lookahead unmasking elicits accurate decoding in diffusion language models
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
fields
cs.LG 3years
2026 3representative citing papers
VRCD prioritizes visually complementary positions during parallel decoding in dMLLMs by measuring attention overlap with the new Visual Redundancy Index, yielding accuracy gains over confidence-based baselines on M^3CoT and MMBench.
Bell-shaped time sampling — drawing the corruption level near t=0.5 — accelerates masked diffusion language model training by up to ~4× without changing the final loss.
citing papers explorer
-
Looped Diffusion Language Models
LoopMDM loops early-middle layers in masked diffusion models to match same-size MDM performance with up to 3.3x fewer training FLOPs and outperform on reasoning tasks by up to 8.5 points on GSM8K.
-
Visual-Redundancy-Controlled Parallel Decoding for Diffusion-Based Multimodal Large Language Models
VRCD prioritizes visually complementary positions during parallel decoding in dMLLMs by measuring attention overlap with the new Visual Redundancy Index, yielding accuracy gains over confidence-based baselines on M^3CoT and MMBench.
-
Understanding and Accelerating the Training of Masked Diffusion Language Models
Bell-shaped time sampling — drawing the corruption level near t=0.5 — accelerates masked diffusion language model training by up to ~4× without changing the final loss.