EXPLORE-Bench: Egocentric Scene Prediction with Long-Horizon Reasoning
This addresses the challenge of long-horizon egocentric reasoning for embodied agents, providing a benchmark for systematic evaluation, though it is incremental in introducing a new task and dataset.
The paper tackles the problem of whether multimodal large language models (MLLMs) can reliably reason about long-term physical consequences of actions from an egocentric viewpoint, and finds a significant performance gap compared to humans, with stepwise reasoning improving performance but at computational cost.
Multimodal large language models (MLLMs) are increasingly considered as a foundation for embodied agents, yet it remains unclear whether they can reliably reason about the long-term physical consequences of actions from an egocentric viewpoint. We study this gap through a new task, Egocentric Scene Prediction with LOng-horizon REasoning: given an initial-scene image and a sequence of atomic action descriptions, a model is asked to predict the final scene after all actions are executed. To enable systematic evaluation, we introduce EXPLORE-Bench, a benchmark curated from real first-person videos spanning diverse scenarios. Each instance pairs long action sequences with structured final-scene annotations, including object categories, visual attributes, and inter-object relations, which supports fine-grained, quantitative assessment. Experiments on a range of proprietary and open-source MLLMs reveal a significant performance gap to humans, indicating that long-horizon egocentric reasoning remains a major challenge. We further analyze test-time scaling via stepwise reasoning and show that decomposing long action sequences can improve performance to some extent, while incurring non-trivial computational overhead. Overall, EXPLORE-Bench provides a principled testbed for measuring and advancing long-horizon reasoning for egocentric embodied perception.