T-MAP: Red-Teaming LLM Agents with Trajectory-aware Evolutionary Search Paper • 2603.22341 • Published 6 days ago • 33
MA-EgoQA: Question Answering over Egocentric Videos from Multiple Embodied Agents Paper • 2603.09827 • Published 17 days ago • 29
MolHIT: Advancing Molecular-Graph Generation with Hierarchical Discrete Diffusion Models Paper • 2602.17602 • Published Feb 19 • 56
view article Article How I Trained Action Chunking Transformer (ACT) on SO-101: My Journey, Gotchas, and Lessons Sep 30, 2025 • 61