How Many Candidates Does a One-Step World Model Need? A Random-Codebook Account of Nearest-of-m Training on Frozen DINO Tokens
cs.LG, cs.CV, cs.RO
Submitted: 2026-11-09
Updated: 2026-11-09
Terminology
Sources
- GE-Act 2.0: Pretraining and Scaling a World-Action Model for Robotic Manipulation
- V-JEPA 2: Self-Supervised Video Models Enable Understanding, Prediction and Planning
- Back to the Features: DINO as a Foundation for Video World Models
- Why the Third Axis Is Freedom
- VFMF: World Modeling by Forecasting Vision Foundation Model Features
- One-Step Prediction Fidelity Is a Poor Proxy for Planning Performance in JEPA World Models
- Conditional Distribution Quantization in Machine Learning
- Sword: Style-Robust World Models as Simulators via Dynamic Latent Bootstrapping for VLA Policy Post-Training
- Improved Mean Flows: On the Challenges of Fastforward Generative Models
- Explorative Modeling: Unlocking a Third Pretraining Axis and End-to-End Generation
- Training Agents Inside of Scalable World Models
- Variational Streaming Flow: Probabilistic Forecasting in Physical Time
- Latent-Foresight: End-to-End Learning Predictable Representations for Latent World Models
- Memorization and Malign Generalization in Conditional Diffusion Models with Random Features
- Beckmann World Models: Direct Terminal-Map Learning for One-Step Action-Conditioned Prediction
- Implicit Maximum Likelihood Estimation
- DriftWorld: Fast World Modeling through Drifting
- ESP: Energy-Score Policy for One-Step Multimodal Action Generation
- Are We Overthinking One-Step Generation?
- Flow Matching in Feature Space for Stochastic World Modeling
Related papers
- A Physics-Regulated Neural Framework for Learning 3D Grain Growth Dynamics
- FlexMoE: Learning One-for-All Multi-Budget Expert Pruning for Flexible MoE Deployment
- OptiMind: Improving Small Language Models for Optimization Modeling through Formulation Guidance and Multi-Turn Learning
- Staying in the Sweet Spot: Boost Reasoning Evolution via Responsive Rollout Difficulty Manipulation
- Examining the Impact of Data and Training Methods on Emergent and Subliminal Misalignment
- Rethinking Black-Box On-Policy Distillation: A Rubric-based Analysis