Search papers, labs, and topics across Lattice.
This paper investigates the phenomenon of hallucination in generative world models, identifying it as a predictable issue concentrated in low-coverage regions of the state-action space. By introducing MMBench2, a comprehensive dataset for visual world modeling, the authors train a 350M-parameter model and uncover three distinct modes of hallucination, each linked to specific stages of the modeling pipeline. They propose a novel coverage-aware sampling technique and utilize hallucination predictors as curiosity rewards, enabling effective adaptation of the model to new environments with minimal additional data.
Hallucination in world models can be predicted and mitigated through targeted data collection strategies, transforming how we approach model training in low-coverage scenarios.
Modern generative world models render increasingly realistic action-controllable futures, yet they frequently hallucinate: rollouts remain visually fluent while drifting from the ground-truth dynamics. We hypothesize that hallucination concentrates in low-coverage regions of the state-action space, where lightweight data-centric signals can both detect it and guide mitigation. To test this, we introduce MMBench2, a 427-hour, 210-task dataset for visual world modeling with ground-truth actions, rewards, and live simulators, and train a 350M-parameter world model on it. We identify three distinct hallucination modes: perceptual, action-marginalized, and scene-diverging -- each anchored to a different stage of the pipeline, and develop three signals that accurately predict where the model will fail. To close coverage gaps at training time, we develop a coverage-aware sampling technique; to close them online, our hallucination predictors serve as curiosity rewards for targeted data collection, yielding a data-efficient finetuning recipe that adapts the pretrained world model to entirely unseen environments with as few as 50 real environment trajectories. Overall, our findings reveal that hallucination in world models is inherently a data coverage issue, and that the same signals used to detect it can also be used for mitigation. An interactive web version of our paper is available at https://www.nicklashansen.com/mmbench2
