人人能懂的AI前沿解读(10.25)

[RO] SuperMimic: Scaling Humanoid Whole-Body Control with One Differentiable World Model
[UC Berkeley]
https://arxiv.org/abs/2610.28480
---
[RO] Agentic World Action Models
[UC Berkeley & University of Pennsylvania]
https://arxiv.org/abs/2610.30136
---
[CL] MemT-1.0: Towards Memory-Native Evolution
[MemTensor Team]
https://arxiv.org/abs/2610.26825
---
[LG] HarnessPro: Recursive Harness Self-Improvement with Pareto Optimization for Performance-Cost Trade-offs
[Tsinghua University & Z.AI]
https://arxiv.org/abs/2610.26770
---
[CL] Large Flow Language Models
[UC Berkeley & Amazon & KAIST]
https://arxiv.org/abs/2610.28559