# NeuroTTT: Bridging Pretraining-Downstream Task Misalignment in EEG Foundation Models via Test-Time Training

arXiv: https://arxiv.org/abs/2509.26301 · HTML: https://mlsanity.com/paper/2509.26301/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 14 of its 48 references that are in the map.
- First posted on arXiv 30 Sep 2025.

## Builds on (most cited first, 14 of 14)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [Masked Autoencoders Are Scalable Vision Learners](https://mlsanity.com/paper/2111.06377/) (2021)
- [Do We Really Need to Access the Source Data? Source Hypothesis Transfer for Unsupervised Domain Adaptation](https://mlsanity.com/paper/2002.08546/) (2020)
- [Large Brain Model for Learning Generic Representations with Tremendous EEG Data in BCI](https://mlsanity.com/paper/2405.18765/) (2024)
- [BENDR: using transformers and a contrastive self-supervised learning task to learn from massive amounts of EEG data](https://mlsanity.com/paper/2101.12037/) (2021)
- [CBraMod: A Criss-Cross Brain Foundation Model for EEG Decoding](https://mlsanity.com/paper/2412.07236/) (2024)
- [Test-Time Training with Self-Supervision for Generalization under Distribution Shifts](https://mlsanity.com/paper/1909.13231/) (2019)
- [Beyond Model Adaptation at Test Time: A Survey](https://mlsanity.com/paper/2411.03687/) (2024)
- [The Brain's Bitter Lesson: Scaling Speech Decoding With Self-Supervised Learning](https://mlsanity.com/paper/2406.04328/) (2024)
- [BIOT: Cross-data Biosignal Learning in the Wild](https://mlsanity.com/paper/2305.10351/) (2023)
- [EEG2Vec: Learning Affective EEG Representations via Variational Autoencoders](https://mlsanity.com/paper/2207.08002/) (2022)
- [SPDIM: Source-Free Unsupervised Conditional and Label Shift Adaptation in EEG](https://mlsanity.com/paper/2411.07249/) (2024)
- [A Simple Review of EEG Foundation Models: Datasets, Advancements and Future Perspectives](https://mlsanity.com/paper/2504.20069/) (2025)
- [Domain-Specific Denoising Diffusion Probabilistic Models for Brain Dynamics](https://mlsanity.com/paper/2305.04200/) (2023)

## Built on by (most cited first, 2 of 2)

- [Extending Pretrained 10-Second ECG Foundation Models to Longer Horizons](https://mlsanity.com/paper/2605.16975/) (2026)
- [Separating personal from population gains when calibrating EEG foundation models for new users](https://mlsanity.com/paper/2609.34801/) (2026)

## Abstract (arXiv)

Large-scale foundation models for EEG signals offer a promising path to generalizable brain-computer interface (BCI) applications, but they often suffer from misalignment between pretraining objectives and downstream tasks, as well as significant cross-subject distribution shifts. This paper addresses these challenges by introducing a two-stage alignment strategy that bridges the gap between generic pretraining and specific EEG decoding tasks. First, we propose NeuroTTT: a domain-specific self-supervised fine-tuning paradigm that augments the foundation model with task-relevant self-supervised objectives, aligning latent representations to important spectral, spatial, and temporal EEG features without requiring additional labeled data. Second, we incorporate test-time training (TTT) at inference, we perform (i) self-supervised test-time training on individual unlabeled test samples and (ii) prediction entropy minimization (Tent), which updates only normalization statistics to continually calibrate the model to each new input on the fly. Our approach, which, to our knowledge, is the first to unify domain-tuned self-supervision with test-time training in large-scale EEG foundation models, yields substantially improved robustness and accuracy across diverse BCI tasks (imagined speech, stress detection, motor imagery). Using CBraMod and LaBraM as backbones, our method pushes their performance to a markedly higher level. Results on three diverse tasks demonstrate that the proposed alignment strategy achieves state-of-the-art performance, outperforming conventional fine-tuning and adaptation methods. Our code is available at https://github.com/wsl2000/NeuroTTT.