# GRIP: Grounded Reasoning via Information-Restricted Premises

arXiv: https://arxiv.org/abs/2608.16776 · HTML: https://mlsanity.com/paper/2608.16776/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in RAG robustness and adaptive retrieval, one of 15 directions in Retrieval-augmented generation.
- Builds on 23 of its 25 references that are in the map.
- First posted on arXiv 17 Aug 2026.

## Placed in

- Retrieval-augmented generation / [RAG robustness and adaptive retrieval](https://mlsanity.com/rag/rag-robustness-and-adaptive-retrieval/)

## Builds on (most cited first, 23 of 23)

- [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://mlsanity.com/paper/2005.11401/) (2020)
- [Dense Passage Retrieval for Open-Domain Question Answering](https://mlsanity.com/paper/2004.04906/) (2020)
- [HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answering](https://mlsanity.com/paper/1809.09600/) (2018)
- [Shortcut Learning in Deep Neural Networks](https://mlsanity.com/paper/2004.07780/) (2020)
- [Self-RAG: Learning to Retrieve, Generate, and Critique through Self-Reflection](https://mlsanity.com/paper/2310.11511/) (2023)
- [Know What You Don't Know: Unanswerable Questions for SQuAD](https://mlsanity.com/paper/1806.03822/) (2018)
- [Constructing A Multi-hop QA Dataset for Comprehensive Evaluation of Reasoning Steps](https://mlsanity.com/paper/2011.01060/) (2020)
- [When Not to Trust Language Models: Investigating Effectiveness of Parametric and Non-Parametric Memories](https://mlsanity.com/paper/2212.10511/) (2022)
- [Measuring and Narrowing the Compositionality Gap in Language Models](https://mlsanity.com/paper/2210.03350/) (2022)
- [How Contextual are Contextualized Word Representations? Comparing the Geometry of BERT, ELMo, and GPT-2 Embeddings](https://mlsanity.com/paper/1909.00512/) (2019)
- [Did Aristotle Use a Laptop? A Question Answering Benchmark with Implicit Reasoning Strategies](https://mlsanity.com/paper/2101.02235/) (2021)
- [Making Retrieval-Augmented Language Models Robust to Irrelevant Context](https://mlsanity.com/paper/2310.01558/) (2023)
- [Trusting Your Evidence: Hallucinate Less with Context-aware Decoding](https://mlsanity.com/paper/2305.14739/) (2023)
- [ProofWriter: Generating Implications, Proofs, and Abductive Statements over Natural Language](https://mlsanity.com/paper/2012.13048/) (2020)
- [MiniCheck: Efficient Fact-Checking of LLMs on Grounding Documents](https://mlsanity.com/paper/2404.10774/) (2024)
- [CLUB: A Contrastive Log-ratio Upper Bound of Mutual Information](https://mlsanity.com/paper/2006.12013/) (2020)
- [Emergence of Invariance and Disentanglement in Deep Representations](https://mlsanity.com/paper/1706.01350/) (2017)
- [Understanding the Limitations of Variational Mutual Information Estimators](https://mlsanity.com/paper/1910.06222/) (2019)
- [Sufficient Context: A New Lens on Retrieval Augmented Generation Systems](https://mlsanity.com/paper/2411.06037/) (2024)
- [Context-DPO: Aligning Language Models for Context-Faithfulness](https://mlsanity.com/paper/2412.15280/) (2024)
- [Beyond Normal: On the Evaluation of Mutual Information Estimators](https://mlsanity.com/paper/2306.11078/) (2023)
- [An Information Bottleneck Perspective for Effective Noise Filtering on Retrieval-Augmented Generation](https://mlsanity.com/paper/2406.01549/) (2024)
- [Accommodate Knowledge Conflicts in Retrieval-augmented LLMs: Towards Robust Response Generation in the Wild](https://mlsanity.com/paper/2504.12982/) (2025)

## Abstract (arXiv)

High-capacity encoders in retrieval-augmented generation (RAG) can let the query dominate the latent state, leaving retrieved evidence functionally irrelevant. We call this failure mode query dominance. To address it, we introduce GRIP (Grounded Reasoning via Information-Restricted Premises), which imposes capacity asymmetry: the decoder keeps full-dimensional access to the query, while retrieved evidence passes through a severe stochastic bottleneck. This forces the evidence channel to encode only the residual information unavailable from the query. Across five reasoning benchmarks, GRIP outperforms strong iterative baselines, cuts a query--latent mutual-information diagnostic by roughly 30× (14.8 → 0.47 bits), and reduces hallucination by 73%. Residual-alignment analysis further shows that the bottleneck output occupies subspaces less aligned with the query than baseline representations.