REVIEW 5 cited by
Reasoning Physical Video Generation with Diffusion Timestep Tokens via Reinforcement Learning
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
Reasoning Physical Video Generation with Diffusion Timestep Tokens via Reinforcement Learning
read the original abstract
Despite recent progress in video generation, producing videos that adhere to physical laws remains a significant challenge. Traditional diffusion-based methods struggle to extrapolate to unseen physical conditions (eg, velocity) due to their reliance on data-driven approximations. To address this, we propose to integrate symbolic reasoning and reinforcement learning to enforce physical consistency in video generation. We first introduce the Diffusion Timestep Tokenizer (DDT), which learns discrete, recursive visual tokens by recovering visual attributes lost during the diffusion process. The recursive visual tokens enable symbolic reasoning by a large language model. Based on it, we propose the Phys-AR framework, which consists of two stages: The first stage uses supervised fine-tuning to transfer symbolic knowledge, while the second stage applies reinforcement learning to optimize the model's reasoning abilities through reward functions based on physical conditions. Our approach allows the model to dynamically adjust and improve the physical properties of generated videos, ensuring adherence to physical laws. Experimental results demonstrate that PhysAR can generate videos that are physically consistent.
Forward citations
Cited by 5 Pith papers
-
Physics Question Scene Graph: Fine-grained Evaluation of Physical Plausibility in Text-to-Video Generation
PQSG is a hierarchical question-graph pipeline for fine-grained physical plausibility evaluation in text-to-video generation that achieves higher correlation with human judgments than prior methods on the FinePhyEval dataset.
-
Incantation: Natural Language as the Action Interface for Multi-Entity Video World Models
Incantation is the first video world model to use per-frame natural language conditioning for simultaneous multi-entity control and concept-level cross-entity transfer in interactive video generation.
-
Incantation: Natural Language as the Action Interface for Multi-Entity Video World Models
Per-frame natural-language action prompts enable simultaneous multi-entity control and cross-entity action transfer in interactive video world models, outperforming discrete action-index interfaces.
-
PhySe-RPO: Physics and Semantics Guided Relative Policy Optimization for Diffusion-Based Surgical Smoke Removal
PhySe-RPO enables diffusion-based surgical smoke removal by converting restoration into a stochastic policy optimized with physics consistency and CLIP semantic rewards under limited supervision.
-
Enhancing Physical Plausibility in Video Generation by Reasoning the Implausibility
A training-free framework uses physics-violating counterfactual prompts and Synchronized Decoupled Guidance to suppress implausible motions in diffusion-based video generation while preserving photorealism.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.