Grounded in Time: A Multi-Source Dataset and Benchmark for Temporal Grounding in Robotic Manipulation

arXiv:2610.04255v1 Announce Type: cross Abstract: Robotic manipulation often requires inferring task-relevant states from past interactions when the current observation alone is insufficient to determine the appropriate action. Despite progress in benchmarking memory-augmented vision-language-action (VLA) models, application-oriented tasks requiring history-dependent semantic inference remain underrepresented. We introduce GiT (Grounded in Time), a dataset and benchmark for grounding manipulation decisions in past events across biolaboratory, household, and industrial scenarios. It includes real-robot and Universal Manipulation Interface (UMI) style demonstrations covering 18 bimanual tasks, together with simulation data and a ManiSkill-based evaluation suite covering nine tasks. Fine-grained subtask annotations and annotated counterfactual task pairs, in which similar current observations require different actions depending on prior events, support policy learning and targeted evaluation of history use. Evaluations of representative end-to-end VLA models in simulation and on selected real-world tasks reveal substantial room for improvement in history-dependent manipulation. The dataset and benchmark are available at the project page.

Related

• RoboIRGBench: Benchmarking Implicit Referential Grounding in Vision-Language-Action Models
• EvoMem-VLA: State-Evolution Memory for Long-Horizon Robot Manipulation
• EventVLA: Event-Driven Visual Evidence Memory for Long-Horizon Vision-Language-Action Policies
• Stress Tests REVEAL Fragile Temporal and Visual Grounding in Video-Language Models
• Fine-Tuning VLAs with Self-Demonstrated Generative Control for Multi-Task Manipulation

Source: arXiv cs.CV | 2026-10-06