# Grounded Revision vs. Prior Injection: Probing Retrieval-Augmented Patent Claim Amendment

arXiv: https://arxiv.org/abs/2609.36550 · HTML: https://mlsanity.com/paper/2609.36550/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Legal and financial RAG, one of 15 directions in Retrieval-augmented generation.
- Builds on 25 of its 35 references that are in the map.
- First posted on arXiv 29 Sep 2026.

## Placed in

- Retrieval-augmented generation / [Legal and financial RAG](https://mlsanity.com/rag/legal-and-financial-rag/)

## Builds on (most cited first, 25 of 25)

- [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://mlsanity.com/paper/2005.11401/) (2020)
- [Evaluating Large Language Models Trained on Code](https://mlsanity.com/paper/2107.03374/) (2021)
- [Lost in the Middle: How Language Models Use Long Contexts](https://mlsanity.com/paper/2307.03172/) (2023)
- [SWE-bench: Can Language Models Resolve Real-World GitHub Issues?](https://mlsanity.com/paper/2310.06770/) (2023)
- [REALM: Retrieval-Augmented Language Model Pre-Training](https://mlsanity.com/paper/2002.08909/) (2020)
- [Self-RAG: Learning to Retrieve, Generate, and Critique through Self-Reflection](https://mlsanity.com/paper/2310.11511/) (2023)
- [BEIR: A Heterogenous Benchmark for Zero-shot Evaluation of Information Retrieval Models](https://mlsanity.com/paper/2104.08663/) (2021)
- [Beyond Accuracy: Behavioral Testing of NLP models with CheckList](https://mlsanity.com/paper/2005.04118/) (2020)
- [Atlas: Few-shot Learning with Retrieval Augmented Language Models](https://mlsanity.com/paper/2208.03299/) (2022)
- [The Power of Noise: Redefining Retrieval for RAG Systems](https://mlsanity.com/paper/2401.14887/) (2024)
- [Towards End-to-End Automation of AI Research](https://mlsanity.com/paper/2606.15497/) (2026)
- [BIGPATENT: A Large-Scale Dataset for Abstractive and Coherent Summarization](https://mlsanity.com/paper/1906.03741/) (2019)
- [Enhancing Noise Robustness of Retrieval-Augmented Language Models with Adaptive Adversarial Training](https://mlsanity.com/paper/2405.20978/) (2024)
- [Preregistering NLP Research](https://mlsanity.com/paper/2103.06944/) (2021)
- [InstructPatentGPT: Training patent language models to follow instructions with human feedback](https://mlsanity.com/paper/2406.16897/) (2024)
- [AutoPatent: A Multi-Agent Framework for Automatic Patent Generation](https://mlsanity.com/paper/2412.09796/) (2024)
- [PaperOrchestra: A Multi-Agent Framework for Automated AI Research Paper Writing](https://mlsanity.com/paper/2604.05018/) (2026)
- [PatentScore: Multi-dimensional Evaluation of LLM-Generated Patent Claims](https://mlsanity.com/paper/2505.19345/) (2025)
- [Towards Better Evaluation for Generated Patent Claims](https://mlsanity.com/paper/2505.11095/) (2025)
- [PANORAMA: A Dataset and Benchmarks Capturing Decision Trails and Rationales in Patent Examination](https://mlsanity.com/paper/2510.24774/) (2025)
- [Patent-CR: A Dataset for Patent Claim Revision](https://mlsanity.com/paper/2412.02549/) (2024)
- [PILOT-Bench: A Benchmark for Legal Reasoning in the Patent Domain with IRAC-Aligned Classification Tasks](https://mlsanity.com/paper/2601.04758/) (2026)
- [PEDANTIC: A Dataset for the Automatic Examination of Definiteness in Patent Claims](https://mlsanity.com/paper/2505.21342/) (2025)
- [PATENTWRITER: A Benchmarking Study for Patent Drafting with LLMs](https://mlsanity.com/paper/2507.22387/) (2025)
- [ToC: Tree-of-Claims Search with Multi-Agent Language Models](https://mlsanity.com/paper/2511.16972/) (2025)

## Abstract (arXiv)

Retrieval-augmented generation is widely used in professional writing, yet whether retrieval grounds revision or merely injects templates is rarely tested where "correct" has a definable meaning. Patent claim amendment supplies that signal: the examiner names the attacked limitation and cites prior art, providing per-case ground truth. We release three artifacts: (i) a corpus of 7,385 USPTO prosecution cases with XML-aligned pre/post claims, rejection, and cited prior art; (ii) a seven-probe battery comparing random and structural-match retrieval as two policies under a fixed prompt scaffold; (iii) a deterministic five-channel metric (C1-C3 and C5 in main, C4 supplementary) requiring no LLM evaluation. Across 9,600 pre-registered calls on four frontier LLMs (Claude Sonnet 4, Claude Haiku 4.5, GPT-5.4, GPT-4o-mini), no tested model exhibits detectable classical prior-injection behavior; retrieval effects are small and direction-inconsistent between random and structural retrieval, and the null is unchanged under a dense (semantic) retriever, across retrieval depths k in {1,3,5,10}, and under a paraphrase-sensitive grounding metric. Revision locality reveals a model-specific difference that the template channel misses. The four-cell taxonomy, which we treat as exploratory, leaves the prior-injector cell unoccupied.