Contrastive World Models: swapping pixel reconstruction for InfoMax boosts robustness
burny_tech · x · 2026-09-28
A new arXiv paper by Bonnie Li, "Contrastive World Models," replaces pixel reconstruction in Dreamer's world model objective with a Deep InfoMax-style lower bound that maximizes mutual information between state-action sequences and local patch features of future observations.
- Motivation: pixel reconstruction lets irrelevant visual details dominate the objective, distracting models from planning-relevant dynamics
- Results: matches Dreamer and a momentum baseline in default settings; substantially outperforms both once moving distractors or natural-video backgrounds are introduced
- Bonus: removing the pixel decoder entirely makes training more efficient
The author argues contrastive infomax objectives are a principled path to world models robust to visual nuisance factors — key for transferring model-based RL agents to the real world.
More from Research
- IMLE-VLA replaces flow matching with single-step cIMLE, 3.67x faster robot actions at 55Hz — petitegeek · 2026-09-28
- Red Queen Bio Uses AI and Wet Labs to Build Medicines Against Natural and Synthetic Viruses — _sholtodouglas · 2026-09-28
- Yacine demos general sim2real pipeline: policy trained in just two minutes — yacineMTB · 2026-09-28
- Claude Spots a Missing 1/2 Coefficient in a Physics Preprint — tak3sh8 · 2026-09-28
- Astra saw through perturbed Zork observations, calling it a 'variant' — tw_killian · 2026-09-28
- FuseReg: Layer-Fusion Regularization Cuts gFID up to 29% in Representation Autoencoders — USC-PSI-Lab · 2026-09-28