runtime-system

REA is a unified runtime that coordinates a deliberative foundation-model brain with fast reactive VLA and whole-body policies under a 'no running policy waits on the brain' rule, achieving 95.9% success on 60 LIBERO-Pro tasks with 25% fewer control steps, plus strong results on RoboTwin 2.0, BEHAVIOR-1K, and a real Unitree G1 humanoid.

DynaResize is a runtime GPU reallocation system that dynamically switches GPUs between Rollout and Training stages during RL-based LLM post-training, improving throughput by 66.5% and reducing execution time by 33% over static configurations.

Tencent Cloud database team open-sourced TencentDB Agent Memory, a runtime system that solves the context degradation problem in long tasks for AI agents, compressing short-term context into the memory system through three-layer backtracking and dynamic compression, and integrating a long-term memory pipeline. This is a landmark attempt for AI agent memory systems moving from 'database' to 'runtime'.