pith. sign in

Dgpo: discovering multiple strategies with diversity-guided policy optimization

1 Pith paper cite this work. Polarity classification is still indexing.

1 Pith paper citing it

fields

cs.CV 1

years

2026 1

verdicts

UNVERDICTED 1

representative citing papers

Video Models Can Reason with Verifiable Rewards

cs.CV · 2026-05-14 · unverdicted · novelty 6.0

VideoRLVR uses SDE-GRPO optimization, dense decomposed rewards, and Early-Step Focus to train video diffusion models on verifiable reasoning tasks, outperforming supervised fine-tuning and other video generators on Maze, FlowFree, and Sokoban.

citing papers explorer

Showing 1 of 1 citing paper.

  • Video Models Can Reason with Verifiable Rewards cs.CV · 2026-05-14 · unverdicted · none · ref 5

    VideoRLVR uses SDE-GRPO optimization, dense decomposed rewards, and Early-Step Focus to train video diffusion models on verifiable reasoning tasks, outperforming supervised fine-tuning and other video generators on Maze, FlowFree, and Sokoban.