#activation-steering

The latest AI and developer news about #activation-steering, with the original source and a short summary.

Study Finds LLM Refusal Behavior Mediated by Single Direction

Researchers demonstrate that removing a single vector from the residual stream of 13 open-source models up to 72B parameters effectively disables refusal mechanisms.

·1·2026.11.08 10:00

Apple Research Finds Fluency Trade-Offs in LLM Conditioning Methods

The study shows activation steering is less effective on instruction-tuned models and often degrades fluency.

·1·2026.09.30 09:00

Study Finds Chat Templates Control LLM Self-Referential Voice, Not Model Weights

Researchers identified a specific activation direction that toggles between disclaimer and experiential self-reports in open-source LLMs.

·2026.09.27 19:00

Apple Researchers Unveil DSAS, a Dynamic Scaling Activation Steering Technique That Intervenes Only When Necessary

Apple researchers have unveiled DSAS, a dynamic scaling Activation Steering technique designed to reduce unnecessary performance degradation.

·1·2026.09.18 09:00

Political censorship revealed inside an LLM's weights

Qwen3.5-9B's political censorship was found to be a small circuit that decomposes into 3 axes.

·1·2026.05.19 09:00

I Think, Therefore I Am — Already

LLM reasoning models make their decision in internal activations before they even begin reasoning in text.

·2026.04.06 09:00

Emotional Concepts in Large Language Models and Their Function

Claude Sonnet 4.5 internally represents emotional concepts that even change its behavior.

·1·2026.04.02 00:00