# Model Tells Itself Where to Attend: Faithfulness Meets Automatic Attention Steering

arXiv: https://arxiv.org/abs/2409.10790 · HTML: https://mlsanity.com/paper/2409.10790/

## Key facts

- Cited by 17 papers in the map (data through 14 Oct 2026).
- Builds on 24 of its 31 references that are in the map.
- First posted on arXiv 16 Sep 2024.

## Builds on (most cited first, 24 of 24)

- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [PyTorch: An Imperative Style, High-Performance Deep Learning Library](https://mlsanity.com/paper/1912.01703/) (2019)
- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks](https://mlsanity.com/paper/1908.10084/) (2019)
- [Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challenge](https://mlsanity.com/paper/1803.05457/) (2018)
- [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)
- [Lost in the Middle: How Language Models Use Long Contexts](https://mlsanity.com/paper/2307.03172/) (2023)
- [Can a Suit of Armor Conduct Electricity? A New Dataset for Open Book Question Answering](https://mlsanity.com/paper/1809.02789/) (2018)
- [GQA: Training Generalized Multi-Query Transformer Models from Multi-Head Checkpoints](https://mlsanity.com/paper/2305.13245/) (2023)
- [REPLUG: Retrieval-Augmented Black-Box Language Models](https://mlsanity.com/paper/2301.12652/) (2023)
- [Trusting Your Evidence: Hallucinate Less with Context-aware Decoding](https://mlsanity.com/paper/2305.14739/) (2023)
- [Faithful Chain-of-Thought Reasoning](https://mlsanity.com/paper/2301.13379/) (2023)
- [Understanding Factuality in Abstractive Summarization with FRANK: A Benchmark for Factuality Metrics](https://mlsanity.com/paper/2104.13346/) (2021)
- [Prompting GPT-3 To Be Reliable](https://mlsanity.com/paper/2210.09150/) (2022)
- [RealTime QA: What's the Answer Right Now?](https://mlsanity.com/paper/2207.13332/) (2022)
- [MRQA 2019 Shared Task: Evaluating Generalization in Reading Comprehension](https://mlsanity.com/paper/1910.09753/) (2019)
- [Large Language Models with Controllable Working Memory](https://mlsanity.com/paper/2211.05110/) (2022)
- [Question Decomposition Improves the Faithfulness of Model-Generated Reasoning](https://mlsanity.com/paper/2307.11768/) (2023)
- [Tell Your Model Where to Attend: Post-hoc Attention Steering for LLMs](https://mlsanity.com/paper/2311.02262/) (2023)
- [What Evidence Do Language Models Find Convincing?](https://mlsanity.com/paper/2402.11782/) (2024)
- [ReEval: Automatic Hallucination Evaluation for Retrieval-Augmented Large Language Models via Transferable Adversarial Attacks](https://mlsanity.com/paper/2310.12516/) (2023)
- [Towards Consistent Natural-Language Explanations via Explanation-Consistency Finetuning](https://mlsanity.com/paper/2401.13986/) (2024)
- [Mitigating Large Language Model Hallucination with Faithful Finetuning](https://mlsanity.com/paper/2406.11267/) (2024)

## Built on by (most cited first, 17 of 17)

- [Steering Language Model Refusal with Sparse Autoencoders](https://mlsanity.com/paper/2411.11296/) (2024)
- [ZeroTuning: Unlocking the Initial Token's Power to Enhance Large Language Models Without Training](https://mlsanity.com/paper/2505.11739/) (2025)
- [LLMSteer: Improving Long-Context LLM Inference by Steering Attention on Reused Contexts](https://mlsanity.com/paper/2411.13009/) (2024)
- [Enhancing Instruction Following of LLMs via Activation Steering with Dynamic Rejection](https://mlsanity.com/paper/2603.06745/) (2026)
- [Attention Reveals More Than Tokens: Training-Free Long-Context Reasoning with Attention-guided Retrieval](https://mlsanity.com/paper/2503.09819/) (2025)
- [CAFE: Retrieval Head-based Coarse-to-Fine Information Seeking to Enhance Multi-Document QA Capability](https://mlsanity.com/paper/2505.10063/) (2025)
- [Graph-Guided Textual Explanation Generation Framework](https://mlsanity.com/paper/2412.12318/) (2024)
- [Static Sandboxes Are Inadequate: Modeling Societal Complexity Requires Open-Ended Co-Evolution in LLM-Based Multi-Agent Simulations](https://mlsanity.com/paper/2510.13982/) (2025)
- [Memory Inception: Latent-Space KV Cache Manipulation for Steering LLMs](https://mlsanity.com/paper/2605.06225/) (2026)
- [Context-Fidelity Boosting: Enhancing Faithful Generation through Watermark-Inspired Decoding](https://mlsanity.com/paper/2604.22335/) (2026)
- [Language Models Can Control Their Own Attention](https://mlsanity.com/paper/2609.02737/) (2026)
- [Mind the Quote: Enabling Quotation-Aware Dialogue in LLMs via Plug-and-Play Modules](https://mlsanity.com/paper/2505.24292/) (2025)
- [Self-Improving is Often Sudden: Enlightenment-style Finetuning for Large-Scale Models](https://mlsanity.com/paper/2607.13395/) (2026)
- [Data-Efficient Adaptation of LLMs via Attention Head Reweighting](https://mlsanity.com/paper/2607.13425/) (2026)
- [Attention is Case-Sensitive](https://mlsanity.com/paper/2608.03711/) (2026)
- [Suppressing Pressure, Amplifying Evidence: Self-Guided Attention Steering to Mitigate Sycophancy and Stubbornness](https://mlsanity.com/paper/2610.04329/) (2026)
- [Lend Me Your Eyes: Instruction-Aware Text Embeddings via Attention Relay](https://mlsanity.com/paper/2610.05564/) (2026)

## Abstract (arXiv)

Large language models (LLMs) have demonstrated remarkable performance across various real-world tasks. However, they often struggle to fully comprehend and effectively utilize their input contexts, resulting in responses that are unfaithful or hallucinated. This difficulty increases for contexts that are long or contain distracting information, which can divert LLMs from fully capturing essential evidence. To address this issue, many works use prompting to help LLMs utilize contextual information more faithfully. For instance, iterative prompting highlights key information in two steps that first ask the LLM to identify important pieces of context and then derive answers accordingly. However, prompting methods are constrained to highlighting key information implicitly in token space, which is often insufficient to fully steer the model's attention. To improve model faithfulness more reliably, we propose AutoPASTA, a method that automatically identifies key contextual information and explicitly highlights it by steering an LLM's attention scores. Like prompting, AutoPASTA is applied at inference time and does not require changing any model parameters. Our experiments on open-book QA demonstrate that AutoPASTA effectively enables models to grasp essential contextual information, leading to substantially improved model faithfulness and performance, e.g., an average improvement of 7.95% for LLAMA3-70B-Instruct. Code will be publicly available at https://github.com/QingruZhang/AutoPASTA .