# Deep Noir: Autonomous Steering Discovery via Architectural Chronometry in Transformer Models

arXiv: https://arxiv.org/abs/2609.20722 · HTML: https://mlsanity.com/paper/2609.20722/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Activation steering methods, one of 13 directions in Mechanistic interpretability.
- Builds on 7 of its 18 references that are in the map.
- First posted on arXiv 17 Sep 2026.

## Placed in

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

## Builds on (most cited first, 7 of 7)

- [Inference-Time Intervention: Eliciting Truthful Answers from a Language Model](https://mlsanity.com/paper/2306.03341/) (2023)
- [Refusal in Language Models Is Mediated by a Single Direction](https://mlsanity.com/paper/2406.11717/) (2024)
- [Open Problems in Mechanistic Interpretability](https://mlsanity.com/paper/2501.16496/) (2025)
- [Semantics-Adaptive Activation Intervention for LLMs via Dynamic Steering Vectors](https://mlsanity.com/paper/2410.12299/) (2024)
- [Taxonomy, Opportunities, and Challenges of Representation Engineering for Large Language Models](https://mlsanity.com/paper/2502.19649/) (2025)
- [Analysing the Safety Pitfalls of Steering Vectors](https://mlsanity.com/paper/2603.24543/) (2026)
- [Transformers Don't Need LayerNorm at Inference Time: Scaling LayerNorm Removal to GPT-2 XL and the Implications for Mechanistic Interpretability](https://mlsanity.com/paper/2507.02559/) (2025)

## Abstract (arXiv)

Activation steering modifies LLM behavior at inference time, but identifying where and how strongly to steer remains manual. We introduce Deep Noir, a framework that uses Logit Lens convergence and causal head-level attribution to autonomously discover optimal steering parameters. Across three scales (1B x 3, 2-3B x 2, and 7-9B x 4), our engine achieves 16.7 percentage-point improvement on spam at 1B (standard deviation 4.7; 39 runs), with gains increasing to 21 to 42 percentage points at 7-9B across four architectures. On SST-2 sentiment, it achieves a 13.1 percentage-point improvement with zero code changes. Mechanistic grounding enables automated discovery of intervention points that generalize across tasks and architectures. On sentiment, RepE without head masking fails to improve over baseline, while Deep Noir improves all models (p less than 0.01). We further show that steering creates a predictable prompt-injection attack surface whose vulnerability increases monotonically with steering magnitude. This finding is relevant to agent systems deploying steered classifiers.