# GRACE: Reinforcement Learning for Grounded Response and Abstention under Contextual Evidence

arXiv: https://arxiv.org/abs/2601.04525 · HTML: https://mlsanity.com/paper/2601.04525/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- Placed in Agentic RAG, agents and memory, one of 15 directions in Retrieval-augmented generation.
- Builds on 27 of its 38 references that are in the map.
- First posted on arXiv 8 Jan 2026.

## Placed in

- Retrieval-augmented generation / [Agentic RAG, agents and memory](https://mlsanity.com/rag/agentic-rag-agents-and-memory/)
- Deep reinforcement learning methods / [RL training of language and agent models](https://mlsanity.com/deeprl/rl-training-of-language-and-agent-models/)

## Builds on (most cited first, 27 of 27)

- [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024)
- [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://mlsanity.com/paper/2005.11401/) (2020)
- [Self-RAG: Learning to Retrieve, Generate, and Critique through Self-Reflection](https://mlsanity.com/paper/2310.11511/) (2023)
- [Search-R1: Training LLMs to Reason and Leverage Search Engines with Reinforcement Learning](https://mlsanity.com/paper/2503.09516/) (2025)
- [Measuring and Narrowing the Compositionality Gap in Language Models](https://mlsanity.com/paper/2210.03350/) (2022)
- [A Dataset of Information-Seeking Questions and Answers Anchored in Research Papers](https://mlsanity.com/paper/2105.03011/) (2021)
- [Contrastive Preference Optimization: Pushing the Boundaries of LLM Performance in Machine Translation](https://mlsanity.com/paper/2401.08417/) (2024)
- [RAFT: Adapting Language Model to Domain Specific RAG](https://mlsanity.com/paper/2403.10131/) (2024)
- [DeepSeek-Prover: Advancing Theorem Proving in LLMs through Large-Scale Synthetic Data](https://mlsanity.com/paper/2405.14333/) (2024)
- [RankRAG: Unifying Context Ranking with Retrieval-Augmented Generation in LLMs](https://mlsanity.com/paper/2407.02485/) (2024)
- [Don't Hallucinate, Abstain: Identifying LLM Knowledge Gaps via Multi-LLM Collaboration](https://mlsanity.com/paper/2402.00367/) (2024)
- [DeepSeekMath-V2: Towards Self-Verifiable Mathematical Reasoning](https://mlsanity.com/paper/2511.22570/) (2025)
- [Search-in-the-Chain: Interactively Enhancing Large Language Models with Search for Knowledge-intensive Tasks](https://mlsanity.com/paper/2304.14732/) (2023)
- [SuRe: Summarizing Retrievals using Answer Candidates for Open-domain QA of LLMs](https://mlsanity.com/paper/2404.13081/) (2024)
- [Multi-hop Question Answering](https://mlsanity.com/paper/2204.09140/) (2022)
- [SimpleDeepSearcher: Deep Information Seeking via Web-Powered Reasoning Trajectory Synthesis](https://mlsanity.com/paper/2505.16834/) (2025)
- [Measuring and Enhancing Trustworthiness of LLMs in RAG through Grounded Attributions and Learning to Refuse](https://mlsanity.com/paper/2409.11242/) (2024)
- [Removal of Hallucination on Hallucination: Debate-Augmented RAG](https://mlsanity.com/paper/2505.18581/) (2025)
- [Ground Every Sentence: Improving Retrieval-Augmented LLMs with Interleaved Reference-Claim Generation](https://mlsanity.com/paper/2407.01796/) (2024)
- [LLM-Based Multi-Hop Question Answering with Knowledge Graph Integration in Evolving Environments](https://mlsanity.com/paper/2408.15903/) (2024)
- [Retrieve-Plan-Generation: An Iterative Planning and Answering Framework for Knowledge-Intensive LLM Generation](https://mlsanity.com/paper/2406.14979/) (2024)
- [Compositional Generalization from Learned Skills via CoT Training: A Theoretical and Structural Analysis for Reasoning](https://mlsanity.com/paper/2502.04667/) (2025)
- [When to Speak, When to Abstain: Contrastive Decoding with Abstention](https://mlsanity.com/paper/2412.12527/) (2024)
- [EviNote-RAG: Enhancing RAG Models via Answer-Supportive Evidence Notes](https://mlsanity.com/paper/2509.00877/) (2025)
- [Debunk the Myth of SFT Generalization](https://mlsanity.com/paper/2510.00237/) (2025)
- [Honest AI: Fine-Tuning "Small" Language Models to Say "I Don't Know", and Reducing Hallucination in RAG](https://mlsanity.com/paper/2410.09699/) (2024)
- [Towards Transparent RAG: Fostering Evidence Traceability in LLM Generation via Reinforcement Learning](https://mlsanity.com/paper/2505.13258/) (2025)

## Built on by (most cited first, 7 of 7)

- [Agent-BRACE: Decoupling Beliefs from Actions in Long-Horizon Tasks via Verbalized State Uncertainty](https://mlsanity.com/paper/2605.11436/) (2026)
- [Do LLM Attribution Metrics Transfer? Auditing Retrieval-Augmented Generation Evaluation Across Datasets and Constructs](https://mlsanity.com/paper/2606.23915/) (2026)
- [MCite-RL: Towards Reliable Multimodal RAG via Citation-enhanced Agentic Reinforcement Learning](https://mlsanity.com/paper/2608.21808/) (2026)
- [Uncertainty Quantification for LLM Agents: A Taxonomy, an Evaluation Protocol, and an Empirical Study](https://mlsanity.com/paper/2609.07395/) (2026)
- [Robust Hierarchical Structures for Agentic Document Analysis](https://mlsanity.com/paper/2609.33322/) (2026)
- [Relevance Is Not Sufficient Evidence: Detecting Evidence Gaps Before Generation in RAG](https://mlsanity.com/paper/2609.37469/) (2026)
- [Training Language Models to Fail Gracefully](https://mlsanity.com/paper/2610.20401/) (2026)

## Abstract (arXiv)

Retrieval-Augmented Generation (RAG) integrates external knowledge to enhance Large Language Models (LLMs), yet systems remain susceptible to two critical flaws: providing correct answers without explicit grounded evidence and producing fabricated responses when the retrieved context is insufficient. While prior research has addressed these issues independently, a unified framework that integrates evidence-based grounding and reliable abstention is currently lacking. In this paper, we propose GRACE, a reinforcement-learning framework that simultaneously mitigates both types of flaws. GRACE employs a data construction method that utilizes heterogeneous retrievers to generate diverse training samples without manual annotation. A multi-stage gated reward function is then employed to train the model to assess evidence sufficiency, extract key supporting evidence, and provide answers or explicitly abstain. Experimental results on two benchmarks demonstrate that GRACE achieves state-of-the-art overall accuracy and strikes a favorable balance between accurate response and rejection, while requiring only 10% of the annotation costs of prior methods. Our code is available at https://github.com/YiboZhao624/Grace..