Vectorize 开源的智能体记忆系统 Hindsight（当日涨星 +4,463、★37,121、MIT）：以世界事实/经验/观察/心智模型四网络替代扁平 RAG，LongMemEval 从同骨干全上下文的 39.0% 拉到 83.6%、最强 91.4%。拆解 TEMPR/CARA 分层、四路检索+RRF+重排、反思持久化机制，附五个落地场景与竞品争议。

Perplexity Research published a new post-training study. It trains a model inside Perplexity Computer on real user sessi…

Multi-turn agents trained with reinforcement learning (RL) receive a single scalar reward per trajectory, which motivate…