HF RL Explorer

Simplify search-based insight type: drop the runtime / backend split

Simplify search-based insight type: drop the runtime / backend split: a task in MiMo-V2.6-RL-oss: Agentic RL Environments (MiMo RL release). While working on the code insights module I keep hitting friction caused by the way search-based insights are typed today. SearchBasedInsight is a union of…

The task

While working on the code insights module I keep hitting friction caused by the way search-based insights are typed today. SearchBasedInsight is a union of SearchRuntimeBasedInsight and SearchBackendBasedInsight, and the two variants have different shapes…

Part of XiaomiMiMo/MiMo-V2.6-RL-oss.