Skip to content
#

uncertainity-estimation

Here are 3 public repositories matching this topic...

Language: All
Filter by language

Budget-constrained RL for human-in-the-loop agent orchestration. A PPO policy decides when to advance, re-retrieve, call a specialist or ask a human, under a hard cap of K reviews per 1,000 queries. Real HotpotQA pipeline, cached-counterfactual replay, audit trail, no synthetic data.

  • Updated Oct 20, 2026
  • Python

Add this topic to your repo

To associate your repository with the uncertainity-estimation topic, visit your repo's landing page and select "manage topics."

Learn more