# Conjecture and Inquiry: Quantifying Software Performance Requirements via Interactive Retrieval-Augmented Preference Elicitation

arXiv: https://arxiv.org/abs/2604.21380 · HTML: https://mlsanity.com/paper/2604.21380/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in RAG pipeline design and evaluation, one of 15 directions in Retrieval-augmented generation.
- Builds on 19 of its 47 references that are in the map.
- First posted on arXiv 23 Apr 2026.

## Placed in

- Retrieval-augmented generation / [RAG pipeline design and evaluation](https://mlsanity.com/rag/rag-pipeline-design-and-evaluation/)

## Builds on (most cited first, 19 of 19)

- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [RoBERTa: A Robustly Optimized BERT Pretraining Approach](https://mlsanity.com/paper/1907.11692/) (2019)
- [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://mlsanity.com/paper/2005.11401/) (2020)
- [Direct Preference Optimization: Your Language Model is Secretly a Reward Model](https://mlsanity.com/paper/2305.18290/) (2023)
- [Representation Learning with Contrastive Predictive Coding](https://mlsanity.com/paper/1807.03748/) (2018)
- [Deep reinforcement learning from human preferences](https://mlsanity.com/paper/1706.03741/) (2017)
- [Joint Embedding of Words and Labels for Text Classification](https://mlsanity.com/paper/1805.04174/) (2018)
- [Astute RAG: Overcoming Imperfect Retrieval Augmentation and Knowledge Conflicts for Large Language Models](https://mlsanity.com/paper/2410.07176/) (2024)
- [WPO: Enhancing RLHF with Weighted Preference Optimization](https://mlsanity.com/paper/2406.11827/) (2024)
- [An Analysis of Fusion Functions for Hybrid Retrieval](https://mlsanity.com/paper/2210.11934/) (2022)
- [Do Performance Aspirations Matter for Guiding Software Configuration Tuning?](https://mlsanity.com/paper/2301.03290/) (2023)
- [MMO: Meta Multi-Objectivization for Software Configuration Tuning](https://mlsanity.com/paper/2112.07303/) (2021)
- [Distilled Lifelong Self-Adaptation for Configurable Systems](https://mlsanity.com/paper/2501.00840/) (2025)
- [PromiseTune: Unveiling Causally Promising and Explainable Configuration Tuning](https://mlsanity.com/paper/2507.05995/) (2025)
- [Faster Configuration Performance Bug Testing with Neural Dual-level Prioritization](https://mlsanity.com/paper/2501.15392/) (2025)
- [Are "Non-functional" Requirements really Non-functional?](https://mlsanity.com/paper/1611.08868/) (2016)
- [Dually Hierarchical Drift Adaptation for Online Configuration Performance Learning](https://mlsanity.com/paper/2507.08730/) (2025)
- [Revealing Domain-Spatiality Patterns for Configuration Tuning: Domain Knowledge Meets Fitness Landscapes](https://mlsanity.com/paper/2603.19897/) (2026)
- [Are Prompts All You Need? Evaluating Prompt-Based Large Language Models (LLM)s for Software Requirements Classification](https://mlsanity.com/paper/2509.13868/) (2025)

## Built on by (most cited first, 1 of 1)

- [LLMSYS-HPOBench: Hyperparameter Optimization Benchmark Suite for Real-World LLM Systems](https://mlsanity.com/paper/2605.08305/) (2026)

## Abstract (arXiv)

Since software performance requirements are documented in natural language, quantifying them into mathematical forms is essential for software engineering. Yet, the vagueness in performance requirements and uncertainty of human cognition have caused highly uncertain ambiguity in the interpretations, rendering their automated quantification an unaddressed and challenging problem. In this paper, we formalize the problem and propose IRAP, an approach that quantifies performance requirements into mathematical functions via interactive retrieval-augmented preference elicitation. IRAP differs from the others in that it explicitly derives from problem-specific knowledge to retrieve and reason the preferences, which also guides the progressive interaction with stakeholders, while reducing the cognitive overhead. Experiment results against 10 state-of-the-art methods on four real-world datasets demonstrate the superiority of IRAP on all cases with up to 40x improvements under as few as five rounds of interactions.