method · first seen 4 Oct 2026
activation steering
Also: latent steering, activation shift, concept injection, activation editing, direction ablation, embedding steering, activation ablation
Modifying a model's internal activations at inference time using steering vectors to change its behaviour without weight updates.
Activation steering changes a model's behaviour at inference time by adding or removing directions in its hidden activations, with no weight updates. The sources show it used as a standard tool rather than introduced. Papers refine when and where to steer (per-token gating, single-prefix steering), compare probing and steering directions, extend it to diffusion language models and diffusion transformers, and use it adversarially to inject beliefs or bias. Small GitHub projects wrap it for local models: neuron explorers for Qwen and GGUF/llama.cpp, and agent tooling that lets models adjust their own activations. Most activity is academic interpretability and safety work.
The work is moving from showing that steering works toward controlling its side effects, timing and robustness. That matters if steering is to serve as a practical alternative to fine-tuning, or if it becomes an attack surface for models with exposed internals.
Analyst note (established), 17 Nov 2026.
Mentions per week
PapersCodeDiscussion
Rank 46 on 14 Nov 2026 · 94 mentions in 28 days · 93 independent groups · papers → code · first seen 6 weeks ago · acceleration +0.69
Earliest sources
- One Token Can Be Enough: Bridging Prompting and Activation Steering with Prefix SteeringPapers · 4 Oct 2026 · Xudong Zhu, Zhihui Zhu
- EmoRSS: Mitigating Emotion-Induced Over-Refusal in Large Language ModelsPapers · 4 Oct 2026 · Shuyi Miao, Yaojin Ma, Chenhang Cui, Xiaohao Liu, Dang Jisheng, Shengda Zhuo, Fe…
- Usage-Modulated Sentiment Representations in Large Language ModelsPapers · 4 Oct 2026 · Hongfei Du, Jiacheng Shi, Yanfu Zhang, Gang Zhou, Ye Gao
- Noise Out, Bias In: Targeted Bias Injection in Diffusion Language Models via Closed-Loop Activation SteeringPapers · 5 Oct 2026 · Sarim Hashmi, Mukul Ranjan, Abdelrahman Elsayed, Muhammad Umer Sheikh, Fahad Sha…
- TrustMI: Causally controlling how assistants trust their usersPapers · 5 Oct 2026 · Théo Lasnier, Romain Froger, Maxence Lasbordes, Djamé Seddah
- From Target Selection to Condition Reuse: Auditing Residual Interventions in Tool-Use ModelsPapers · 5 Oct 2026 · Tianda Sun, Dimitar Kazakov
Latest sources
- Abstention Direction in Large Language Models: Mechanistic Interpretability of Abstention Across Query Ambiguity, Knowledge Limits, and Safety AlignmentPapers · 16 Nov 2026 · Koki Itai, Jun Hirako, Hiroaki Yamagiwa
- Refusal Register is Causally Steerable: Decomposing the Refusal Direction into Category-Aligned Components in Gemma-2-9B-ITPapers · 16 Nov 2026 · Janhavi Khindkar, Nataliia Povarova
- ShutongLiang/Neural-State-OSCode · 14 Nov 2026 · ShutongLiang · 84 stars
- Do Probing and Steering Favor the Same Directions Within and Across Language Models?Papers · 14 Nov 2026 · Afshin Sadeghi, Daniel Becker
- Few-Shot Visual Concept Extraction for Steering Diffusion TransformersPapers · 14 Nov 2026 · Nabyl Quignon, Antitza Dantcheva
- The Geometry of Conceptors in Neural-Activation SpacesPapers · 14 Nov 2026 · Ilias Triantafyllopoulos, Young Min Cho, Ren Tao, Miranda Muqing Miao, Sunny Rai…
- Can Mechanistic Interpretability Help Explain LLM Agent Failures? A Case Study on Web Element HallucinationPapers · 13 Nov 2026 · Mohamed Aghzal, Mor Geva, Gregory J. Stein, Ziyu Yao
- Right Token, Right Time: From Per-Token Gating to Trajectory-Aware Activation SteeringPapers · 13 Nov 2026 · Nima Eshraghi, Lovedeep Gondara, Sagarika Suresh, Xiaotong Xu, Jithin Pradeep, F…
- Reasoning Models Confabulate to Defend Injected Beliefs About Their Own RequestsPapers · 13 Nov 2026 · Yang Ou
- AmbiSteer: Selectively Eliciting Ambiguity-Aware Responses in Vision-Language ModelsPapers · 13 Nov 2026 · Monica Riedler, Samuele Antonelli, Barbara Plank, Raffaella Bernardi
- Antidistillation is Cheap: Lightweight Distillation-Resistance via Stylistic CorruptionPapers · 13 Nov 2026 · Anh Tu Ngo, Anupam Chattopadhyay
- Activation Steering the Nan-Shan in TyphoonPapers · 13 Nov 2026 · Megan Wei, Damon Pham, John Cayley, Ellie Pavlick