# Mechanistic Interpretability Reveals Shared Causal Subspaces in Brain-to-Speech Decoders

arXiv: https://arxiv.org/abs/2609.31992 · HTML: https://mlsanity.com/paper/2609.31992/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Circuits and mechanistic analysis, one of 13 directions in Mechanistic interpretability.
- Builds on 11 of its 21 references that are in the map.
- First posted on arXiv 25 Sep 2026.

## Placed in

- Mechanistic interpretability / [Circuits and mechanistic analysis](https://mlsanity.com/mechinterp/circuits-and-mechanistic-analysis/)

## Builds on (most cited first, 11 of 11)

- [Learning Phrase Representations using RNN Encoder-Decoder for Statistical Machine Translation](https://mlsanity.com/paper/1406.1078/) (2014)
- [HiFi-GAN: Generative Adversarial Networks for Efficient and High Fidelity Speech Synthesis](https://mlsanity.com/paper/2010.05646/) (2020)
- [Towards Automated Circuit Discovery for Mechanistic Interpretability](https://mlsanity.com/paper/2304.14997/) (2023)
- [How to use and interpret activation patching](https://mlsanity.com/paper/2404.15255/) (2024)
- [Everything, Everywhere, All at Once: Is Mechanistic Interpretability Identifiable?](https://mlsanity.com/paper/2502.20914/) (2025)
- [Beyond Transcription: Mechanistic Interpretability in ASR](https://mlsanity.com/paper/2508.15882/) (2025)
- [Unboxing the Black Box: Mechanistic Interpretability for Algorithmic Understanding of Neural Networks](https://mlsanity.com/paper/2511.19265/) (2025)
- [Zero-Shot Imagined Speech Decoding via Imagined-to-Listened MEG Mapping](https://mlsanity.com/paper/2605.08075/) (2026)
- [Interpretable Embeddings of Speech Enhance and Explain Brain Encoding Performance of Audio Models](https://mlsanity.com/paper/2507.16080/) (2025)
- [Relating the Neural Representations of Vocalized, Mimed, and Imagined Speech](https://mlsanity.com/paper/2602.22597/) (2026)
- [Transfer Learning for Covert Speech Classification Using EEG Hilbert Envelope and Temporal Fine Structure](https://mlsanity.com/paper/2502.04132/) (2025)

## Abstract (arXiv)

Decoding covert speech, such as mimed or imagined, from brain activity is harder than decoding vocalized speech. Cross-modal transfer, where information from one speech form helps decode another, is a promising remedy; yet how a decoder internally represents and processes brain activity from different speech forms remains unclear. In this work, we ask: which internal neurons of a decoder carry cross-modal information, and are these neurons shared across different speech forms? To answer these questions, we leverage mechanistic interpretability, using recordings of the same sentences in vocalized, mimed, and imagined input pairs for activation patching. We insert the decoder's internal activity for a sentence in one condition into its processing of the same sentence in another and measure the change in decoding accuracy. We find that no single neuron drives this benefit; instead, it arises from small groups of neurons, with vocalized speech as the most useful source. These groups are largely condition-specific in the early stage of the decoder but overlap in the later stage. These findings point toward more data-efficient covert speech decoders through training objectives that encourage shared later-stage representations learned mainly from vocalized data.