# Only Ask What You Don't Know: Grounded Delta Planning for Efficient Multi-step RAG

Source: [Only Ask What You Don't Know: Grounded Delta Planning for Efficient Multi-step RAG](<https://www.sota2.com/research/paper/6a3a1362206542251d7c7406>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-06-24T05:00:22.884Z

Authors: Wei-Chieh Chou, Xuanjun Chen, Jian-Ren Lin, Claire Lin, Hung-yi Lee, Jyh-Shing Roger Jang

Published: 2026-06-21T21:46:41.000Z

## Abstract

Multi-hop question answering remains challenging for Retrieval-Augmented Generation (RAG) because existing approaches either propagate errors across iterative retrieval rounds or over-generate reasoning steps, increasing cost without improving accuracy. We propose Grounded Delta Planning RAG (GDP-RAG), a plan-based framework that targets only the information delta based on three simple design choices: (1) preliminary retrieval to ground planning before execution, (2) a gap-conditioned planning prompt that asks only for missing information, and (3) a skeletal trajectory that pairs each subquery with a Thought capturing evidence from preliminary retrieval and carrying it through to the final answer. GDP-RAG focuses computation on unresolved gaps, yielding concise, reliable reasoning trajectories. Extensive experiments on HotpotQA, 2WikiMultiHopQA, and MuSiQue show that GDP-RAG achieves the highest accuracy (60.63%) among all compared systems while maintaining a cost-of-pass of 0.51, 22% lower than PAR-RAG (0.65) and 68% lower than KnowTrace (1.57), with no method achieving both higher accuracy and lower cost.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Multi-hop Question Answering on HotpotQA l1](<https://www.sota2.com/research/sota/multi-hop-question-answering-on-hotpotqa-l1.md>): Multi-hop Question Answering; HotpotQA l1; Cost per Pass: 0.25.

- [Multi-hop Question Answering on 2WikiMultiHopQA l2](<https://www.sota2.com/research/sota/multi-hop-question-answering-on-2wikimultihopqa-l2.md>): Multi-hop Question Answering; 2WikiMultiHopQA l2; Cost per Pass: 0.3.

- [Multi-hop Question Answering on MusiQue l3](<https://www.sota2.com/research/sota/multi-hop-question-answering-on-musique-l3.md>): Multi-hop Question Answering; MusiQue l3; Cost: 0.98.

- [Multi-hop Question Answering on Average (HotpotQA, 2WikiMultiHopQA, MusiQue) (Cost/Efficiency)](<https://www.sota2.com/research/sota/multi-hop-question-answering-on-average-hotpotqa-2wikimultihopqa-musique-cost-efficiency.md>): Multi-hop Question Answering; HotpotQA, 2WikiMultiHopQA, and MusiQue Average all; Cost per Pass: 0.51.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2606.22681>)

[PDF](<https://arxiv.org/pdf/2606.22681.pdf>)