[KongchangAI]
ConceptRPL

Retrieval Program Learning

A framework in which a model autonomously writes executable retrieval programs through reinforcement learning, composing retrieval primitives into programs with local state and control flow, thereby decoupling the scope of retrieval exploration from context occupancy.

Timeline (last 90 days)

Oct 20

The RPL (Retrieval Program Learning) framework lets a policy model compose retrieval primitives into executable programs with local state and control flow, and optimizes these programs via reinforcement learning.

Unverified50%

All Facts (1)

Unverified

The RPL (Retrieval Program Learning) framework lets a policy model compose retrieval primitives into executable programs with local state and control flow, and optimizes these programs via reinforcement learning.

50%

Source Articles

Reinforcement Learning Lets AI Write Its Own Retrieval Programs, 4B Model Outperforms 8B BaselinesOct 20