RESEARCH

WorldLines: Benchmarking and Modeling Long-Horizon Stateful Embodied Agents

ArXiv cs.AI · Thu, 18 Jun 2026 04:00:00 GMT

arXiv:2606.18847v1 Announce Type: new Abstract: To assist humans over extended periods in real homes, embodied agents must remember user routines, world states, and past interactions. Existing long-term memory benchmarks mainly evaluate language-centric retrieval and question ans

Read original source Discuss with A.S.I.S