# LinEAS: End-to-end Learning of Activation Steering with a Distributional Loss

arXiv: https://arxiv.org/abs/2503.10679 · HTML: https://mlsanity.com/paper/2503.10679/

## Key facts

- Cited by 14 papers in the map (data through 14 Oct 2026).
- Placed in Activation steering methods, one of 13 directions in Mechanistic interpretability.
- Builds on 29 of its 47 references that are in the map.
- First posted on arXiv 11 Mar 2025.

## Placed in

- Mechanistic interpretability / [Activation steering methods](https://mlsanity.com/mechinterp/activation-steering-methods/)

## Builds on (most cited first, 29 of 29)

- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017)
- [LoRA: Low-Rank Adaptation of Large Language Models](https://mlsanity.com/paper/2106.09685/) (2021)
- [DINOv2: Learning Robust Visual Features without Supervision](https://mlsanity.com/paper/2304.07193/) (2023)
- [Measuring Massive Multitask Language Understanding](https://mlsanity.com/paper/2009.03300/) (2020)
- [Classifier-Free Diffusion Guidance](https://mlsanity.com/paper/2207.12598/) (2022)
- [SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesis](https://mlsanity.com/paper/2307.01952/) (2023)
- [Finetuned Language Models Are Zero-Shot Learners](https://mlsanity.com/paper/2109.01652/) (2021)
- [TruthfulQA: Measuring How Models Mimic Human Falsehoods](https://mlsanity.com/paper/2109.07958/) (2021)
- [CLIPScore: A Reference-free Evaluation Metric for Image Captioning](https://mlsanity.com/paper/2104.08718/) (2021)
- [Representation Engineering: A Top-Down Approach to AI Transparency](https://mlsanity.com/paper/2310.01405/) (2023)
- [Inference-Time Intervention: Eliciting Truthful Answers from a Language Model](https://mlsanity.com/paper/2306.03341/) (2023)
- [RealToxicityPrompts: Evaluating Neural Toxic Degeneration in Language Models](https://mlsanity.com/paper/2009.11462/) (2020)
- [BitFit: Simple Parameter-efficient Fine-tuning for Transformer-based Masked Language-models](https://mlsanity.com/paper/2106.10199/) (2021)
- [Steering Llama 2 via Contrastive Activation Addition](https://mlsanity.com/paper/2312.06681/) (2023)
- [Improved Distribution Matching Distillation for Fast Image Synthesis](https://mlsanity.com/paper/2405.14867/) (2024)
- [Learning Generative Models with Sinkhorn Divergences](https://mlsanity.com/paper/1706.00292/) (2017)
- [A General Framework for Inference-time Scaling and Steering of Diffusion Models](https://mlsanity.com/paper/2501.06848/) (2025)
- [ReFT: Representation Finetuning for Language Models](https://mlsanity.com/paper/2404.03592/) (2024)
- [TFG: Unified Training-Free Guidance for Diffusion Models](https://mlsanity.com/paper/2409.15761/) (2024)
- [Residual Connections Encourage Iterative Inference](https://mlsanity.com/paper/1710.04773/) (2017)
- [Controlling Language and Diffusion Models by Transporting Activations](https://mlsanity.com/paper/2410.23054/) (2024)
- [Gradient Guidance for Diffusion Models: An Optimization Perspective](https://mlsanity.com/paper/2404.14743/) (2024)
- [LoFiT: Localized Fine-tuning on LLM Representations](https://mlsanity.com/paper/2406.01563/) (2024)
- [Continuous, Subject-Specific Attribute Control in T2I Models by Identifying Semantic Directions](https://mlsanity.com/paper/2403.17064/) (2024)
- [Whispering Experts: Neural Interventions for Toxicity Mitigation in Language Models](https://mlsanity.com/paper/2407.12824/) (2024)
- [Stylus: Automatic Adapter Selection for Diffusion Models](https://mlsanity.com/paper/2404.18928/) (2024)
- [Self-conditioning pre-trained language models](https://mlsanity.com/paper/2110.02802/) (2021)
- [Realistic Evaluation of Toxicity in Large Language Models](https://mlsanity.com/paper/2405.10659/) (2024)

## Built on by (most cited first, 14 of 14)

- [Steering at the Source: Style Modulation Heads for Robust Persona Control](https://mlsanity.com/paper/2603.13249/) (2026)
- [TADA! Tuning Audio Diffusion Models through Activation Steering](https://mlsanity.com/paper/2602.11910/) (2026)
- [Activation Steering of Video Generation Models via Reduced-Order Linear Optimal Control](https://mlsanity.com/paper/2606.04775/) (2026)
- [Steering Robustness into World Action Models via Mechanistic Interpretability and Optimal Control](https://mlsanity.com/paper/2607.14943/) (2026)
- [Exploring Diverse Generation Paths via Inference-time Stiefel Activation Steering](https://mlsanity.com/paper/2601.22010/) (2026)
- [ELROND: Exploring and decomposing intrinsic capabilities of diffusion models](https://mlsanity.com/paper/2602.10216/) (2026)
- [GAPS: Dimension-Level Gates for Conditional Activation Steering](https://mlsanity.com/paper/2609.01878/) (2026)
- [HyperTransport: Amortized Conditioning of T2I Generative Models](https://mlsanity.com/paper/2605.08254/) (2026)
- [Steered Generation via Gradient-Based Optimization on Sparse Query Features](https://mlsanity.com/paper/2605.23040/) (2026)
- [TimpaTeks: Automatic In-place Text Sequence Modification via Diffusion Language Model Steering](https://mlsanity.com/paper/2606.08408/) (2026)
- [IDEEA: training-free Input-Dependent stEEring via Activation cluster matching](https://mlsanity.com/paper/2609.02089/) (2026)
- [LocUS: Head Selection and Subspace Projection for Targeted Activation Steering](https://mlsanity.com/paper/2609.31122/) (2026)
- [When a Conceptor Complement Becomes Gain: Operator Geometry and Behavioral Attribution in Language Models](https://mlsanity.com/paper/2610.16325/) (2026)
- [On the Limits of Representation-Based Test-Time Adaptation for Multimodal LLMs](https://mlsanity.com/paper/2610.19892/) (2026)

## Abstract (arXiv)

The growing use of generative models in daily life calls for efficient mechanisms to control their generation, to e.g., produce safe content or provide users with tools to explore style changes. Ideally, such mechanisms should require low volume of unpaired data (i.e., without explicit preference), and should be cheap, both at train and inference time, while preserving output quality. Recent research has shown that such mechanisms can be obtained by intervening exclusively on model activations, with the goal of correcting distributional differences between activations seen when using prompts from a source vs. a target set (e.g., toxic and non-toxic sentences). While cheap, these fast methods are inherently crude: their maps are tuned locally, not accounting for their impact on downstream layers, resulting in interventions that cause unintended shifts when used out-of-sample. We propose in this work linear end-to-end activation steering (LinEAS), an approach trained with a global loss that accounts simultaneously for all layer-wise distributional shifts. In addition to being more robust, the loss used to train LinEAS can be regularized with sparsifying norms, which can automatically carry out neuron selection. LinEAS only requires a handful of unpaired samples to be effective, and beats similar baselines on toxicity mitigation in language models, becoming competitive with oracle-dependent methods that have access to strong supervision. LinEAS is modality-agnostic and we empirically find that it outperforms existing activation steering methods at mitigating and including new concepts at the output of single-step text-to-image generation models.