Dimple is a 7B vision-language model that generates answers via discrete diffusion, trained with an autoregressive-then-diffusion recipe, and it matches or beats LLaVA-NEXT on several benchmarks while enabling parallel decoding.
Eagle 2.5: Boosting long-context post-training for frontier vision-language models, 2025
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
method 1
citation-polarity summary
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1roles
method 1polarities
use method 1representative citing papers
citing papers explorer
-
Dimple: Discrete Diffusion Multimodal Large Language Model with Parallel Decoding
Dimple is a 7B vision-language model that generates answers via discrete diffusion, trained with an autoregressive-then-diffusion recipe, and it matches or beats LLaVA-NEXT on several benchmarks while enabling parallel decoding.