RLRAG-Cost-Aware-Retrieval-Depth-Control-for-Production-RAG/CITATION.cff at main · ORION2809/RLRAG-Cost-Aware-Retrieval-Depth-Control-for-Production-RAG · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
cff-version: 1.2.0
message: "If you use this software, please cite it as below."
title: "Adaptive Cost-Aware RAG Controller"
authors:
  - family-names: Suvarna
    given-names: Shreyas
type: software
date-released: 2025-06-01
url: "https://github.com/ORION2809/RLRAG-Cost-Aware-Retrieval-Depth-Control-for-Production-RAG"
license: MIT
keywords:
  - reinforcement-learning
  - retrieval-augmented-generation
  - cost-optimization
  - adaptive-retrieval
  - NLP
abstract: >-
  An RL-based controller that dynamically selects retrieval depth in
  RAG systems, balancing answer quality against inference cost.
  Achieves 28.8% token reduction vs static k=4 baseline with
  minimal F1 loss on Natural Questions.