RAG Retrieval Evaluation


This evaluation tests the performance of our Retrieval-Augmented Generation pipeline for answering questions about reported issues. Explore how different models, top-K values, and prompts affect retrieval quality. The charts below show recall@k and precision@k for each test run.


← Back to Dashboard

DateRecall@kPrecision@kDetailsCommit hash
20250930_2318JSON17a65130febd4303d59bad485288be1b8e7a3175
20250930_1656JSONba5c5de85fa8aa36fcce0f474f5a2d1962eca461
20250929_2317JSON902149ec3a33cccc6cd4dc40c664275a9e9601ad
20250928_1347JSONd60dda7a1ed5544ef034887073533227ccaf9810
20250928_1515JSON2e1a820e69471116bb6c4a1e18754bae4e735636
20250928_1515JSON2e1a820e69471116bb6c4a1e18754bae4e735636
20250928_0827JSON8b8d329194a3319c928ed691e141ab220ffbc5a0
20250928_0827JSON8b8d329194a3319c928ed691e141ab220ffbc5a0
20250928_0827JSON8b8d329194a3319c928ed691e141ab220ffbc5a0
20250926_1218JSONunknown

Report updated: 2025-09-30 23:25