.png){kind=icon}
.png){kind=icon}
Return
ALICE: Autonomous Lifelong Intelligence Framework for Cross-Embodiment via Continuous Internal States Feedback Mechanism
.png){kind=icon}
OA
AI
Z
Z
Y
N
Y
X
DOI：10.1016/j.future.2026.108730.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• ALICE is a modular cross-embodiment framework that decouples perception, memory, decision, and execution for controlled experimentation. • A unified capability registry and action abstraction support coherent operation across heterogeneous tools and embodiments. • CAB is a log-first, diagnostic, non-comparative benchmark that reports interpretable long-horizon behavior signals from runtime traces. • An empirical case study shows structured action sequencing and long-tailed thinking-chain dynamics in continuous operation.
Keywords:
Embodied AI
Cross-embodiment agents
Continuous autonomous agents
Agent evaluation
Modular agent frameworks
Hierarchical task decomposition
Journal
F
IF:
6.1
Papers:
6.8K
Citations:
2.3W
Organization
Cited Papers