# Necessary, Decodable and Reversible, Yet Not Transferable: A Stress Test for Attention-Head Role Claims

arXiv: https://arxiv.org/abs/2606.08292 · HTML: https://mlsanity.com/paper/2606.08292/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 24 of its 27 references that are in the map.
- First posted on arXiv 6 Jun 2026.

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

- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Qwen2.5 Technical Report](https://mlsanity.com/paper/2412.15115/) (2024)
- [Mistral 7B](https://mlsanity.com/paper/2310.06825/) (2023)
- [Locating and Editing Factual Associations in GPT](https://mlsanity.com/paper/2202.05262/) (2022)
- [Interpretability in the Wild: a Circuit for Indirect Object Identification in GPT-2 small](https://mlsanity.com/paper/2211.00593/) (2022)
- [Toy Models of Superposition](https://mlsanity.com/paper/2209.10652/) (2022)
- [Probing Classifiers: Promises, Shortcomings, and Advances](https://mlsanity.com/paper/2102.12452/) (2021)
- [Discovering Latent Knowledge in Language Models Without Supervision](https://mlsanity.com/paper/2212.03827/) (2022)
- [Towards Automated Circuit Discovery for Mechanistic Interpretability](https://mlsanity.com/paper/2304.14997/) (2023)
- [Designing and Interpreting Probes with Control Tasks](https://mlsanity.com/paper/1909.03368/) (2019)
- [In-Context Learning Creates Task Vectors](https://mlsanity.com/paper/2310.15916/) (2023)
- [Function Vectors in Large Language Models](https://mlsanity.com/paper/2310.15213/) (2023)
- [How to use and interpret activation patching](https://mlsanity.com/paper/2404.15255/) (2024)
- [Linearity of Relation Decoding in Transformer Language Models](https://mlsanity.com/paper/2308.09124/) (2023)
- [Open Problems in Mechanistic Interpretability](https://mlsanity.com/paper/2501.16496/) (2025)
- [Is This the Subspace You Are Looking for? An Interpretability Illusion for Subspace Activation Patching](https://mlsanity.com/paper/2311.17030/) (2023)
- [Copy Suppression: Comprehensively Understanding an Attention Head](https://mlsanity.com/paper/2310.04625/) (2023)
- [Which Attention Heads Matter for In-Context Learning?](https://mlsanity.com/paper/2502.14010/) (2025)
- [The Super Weight in Large Language Models](https://mlsanity.com/paper/2411.07191/) (2024)
- [Interpretability Illusions in the Generalization of Simplified Models](https://mlsanity.com/paper/2312.03656/) (2023)
- [Mechanistic Interpretability as Statistical Estimation: A Variance Analysis](https://mlsanity.com/paper/2510.00845/) (2025)
- [Causality neq Invariance: Function and Concept Vectors in LLMs](https://mlsanity.com/paper/2602.22424/) (2026)
- [Beyond Components: Singular Vector-Based Interpretability of Transformer Circuits](https://mlsanity.com/paper/2511.20273/) (2025)
- [Make Mechanistic Interpretability Auditable: A Call to Develop Guidelines via Continuous Collaborative Reviewing](https://mlsanity.com/paper/2606.00033/) (2026)

## Abstract (arXiv)

Mechanistic studies often assign a component a role when removing it damages a behavior, its activation linearly encodes task information, and restoring that activation repairs the damage. We test the stronger implication: can the same activation carry the requested computation into another prompt? In the roughly 8 of 15 model-family cells receiving our full matched-control transfer assay, covering three 7-8B instruction-tuned models and 5 single-step computation families, tested attention-head states did not produce clean computation transfer. Same-answer and matched-context controls instead exposed inert or broad effects. Positive controls bound this result rather than eliminating its caveats: the instrument recovers known function vectors when writing into an underspecified prompt, and recovers a compact override-capable residual state in one of three controlled-model seeds. Neither control proves sensitivity to the object carried by real-model attention during prompt override. On two composed forms in Qwen, a strongly decodable operation direction also failed to steer the generated answer under passing gates; the corresponding cross-model test was invalid. These results show that necessity, decodability, same-prompt repair, and cross-prompt transfer are separable evidence. They do not establish that a portable selection state is absent from untested components or settings.