# Spurious reconstruction from brain activity

arXiv: https://arxiv.org/abs/2405.10078 · HTML: https://mlsanity.com/paper/2405.10078/

## Key facts

- Cited by 15 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 16 May 2024.

## Built on by (most cited first, 15 of 15)

- [Scaling laws for decoding images from brain activity](https://mlsanity.com/paper/2501.15322/) (2025)
- [NSD-Imagery: A benchmark dataset for extending fMRI vision decoding methods to mental imagery](https://mlsanity.com/paper/2506.06898/) (2025)
- [Dynadiff: Single-stage Decoding of Images from Continuously Evolving fMRI](https://mlsanity.com/paper/2505.14556/) (2025)
- [Learning Interpretable Representations Leads to Semantically Faithful EEG-to-Text Generation](https://mlsanity.com/paper/2505.17099/) (2025)
- [Scaling Vision Transformers for Functional MRI with Flat Maps](https://mlsanity.com/paper/2510.13768/) (2025)
- [Meta-Learning an In-Context Transformer Model of Human Higher Visual Cortex](https://mlsanity.com/paper/2505.15813/) (2025)
- [Towards Interpretable Visual Decoding with Attention to Brain Representations](https://mlsanity.com/paper/2509.23566/) (2025)
- [Rethinking Brain Decoding with CLIP: The Role of Adversarial Robustness](https://mlsanity.com/paper/2607.03165/) (2026)
- [Deep Neural Encoder-Decoder Model to Relate fMRI Brain Activity with Naturalistic Stimuli](https://mlsanity.com/paper/2507.12009/) (2025)
- [MEIcoder: Decoding Visual Stimuli from Neural Activity by Leveraging Most Exciting Inputs](https://mlsanity.com/paper/2510.20762/) (2025)
- [Simple Models, Rich Representations: Visual Decoding from Primate Intracortical Neural Signals](https://mlsanity.com/paper/2601.11108/) (2026)
- [The Umwelt Representation Hypothesis: Rethinking Universality](https://mlsanity.com/paper/2604.17960/) (2026)
- [MIRAGE: Robust multi-modal architectures translate fMRI-to-image models from vision to mental imagery](https://mlsanity.com/paper/2605.17198/) (2026)
- [Real-time Reconstruction of Human Visual Perception from fMRI](https://mlsanity.com/paper/2607.22753/) (2026)
- [AI-Driven Neural Surrogates for In Silico Design of Cognitive-Affective Neuromodulation Targets](https://mlsanity.com/paper/2609.27729/) (2026)

## Abstract (arXiv)

Advances in brain decoding, particularly visual image reconstruction, have sparked discussions about the societal implications and ethical considerations of neurotechnology. As these methods aim to recover visual experiences from brain activity and achieve prediction beyond training samples (zero-shot prediction), it is crucial to assess their capabilities and limitations to inform public expectations and regulations. Our case study of recent text-guided reconstruction methods, which leverage a large-scale dataset (Natural Scene Dataset, NSD) and text-to-image diffusion models, reveals limitations in their generalizability. We found poor performance when applying these methods to a different dataset designed to prevent category overlaps between training and test sets. UMAP visualization of the text features with NSD images showed a limited diversity of semantic and visual clusters, with overlap between training and test sets. Formal analysis and simulations demonstrated that clustered training samples can lead to "output dimension collapse," restricting predictable output feature dimensions. Simulations further showed that diversifying the training set improved generalizability. However, text features alone are insufficient for mapping to the visual space. We argue that recent realistic reconstructions may primarily be a blend of classification into trained categories and generation of inauthentic images through text-to-image diffusion (hallucination). Diverse datasets and compositional representations spanning the image space are essential for genuine zero-shot prediction. Interdisciplinary discussions grounded in understanding the current capabilities and limitations, as well as ethical considerations, of the technology are crucial for its responsible development.