The Semantic Bottleneck: Leveraging Semantic Representations for Non-Invasive Speech Decoding
Summary
This paper introduces Brain2Semantics2Text, a non-invasive speech decoding method that maps MEG responses to semantic embeddings to reconstruct sentence-level text without word-level alignment.
View Cached Full Text
Cached at: 09/10/26, 02:16 PM
Paper page - The Semantic Bottleneck: Leveraging Semantic Representations for Non-Invasive Speech Decoding
Source: https://huggingface.co/papers/2609.10296
Abstract
A non-invasive brain decoding approach maps MEG responses to semantic embeddings to reconstruct sentence-level text without requiring word-level alignment.
Non-invasive speech decodingremains constrained by the low signal-to-noise ratio of neural recordings, which makes fine-grained reconstruction of phonemes or individual words difficult. Motivated by neuroscientific evidence that high-level semantic representations are distributed across cortical regions and evolve over slower temporal scales, we hypothesize that semantic content may provide a more suitable target for non-invasive decoding than low-level acoustic or lexical features. We introduceBrain2Semantics2Text, a method that reconstructs text through an intermediatesemantic embedding space. Our model maps sentence-levelMEGresponses into a semantic manifold and then inverts the predicted embeddings into natural language. This semantic bottleneck enables recovery of high-level meaning without word-level alignment. We describe the core principles of the approach, its implementation, and the strategies used to mitigate the challenges of learning a reliableneural-to-semantic mapping. Finally, we compare against prior non-invasive Brain2Text methods and show improved sentence-level results.
View arXiv pageView PDFAdd to collection
Get this paper in your agent:
hf papers read 2609\.10296
Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash
Models citing this paper0
No model linking this paper
Cite arxiv.org/abs/2609.10296 in a model README.md to link it from this page.
Datasets citing this paper0
No dataset linking this paper
Cite arxiv.org/abs/2609.10296 in a dataset README.md to link it from this page.
Spaces citing this paper0
No Space linking this paper
Cite arxiv.org/abs/2609.10296 in a Space README.md to link it from this page.
Collections including this paper0
No Collection including this paper
Add this paper to acollectionto link it from this page.
Similar Articles
Brain-CLIPLM: Decoding Compressed Semantic Representations in EEG for Language Reconstruction
Researchers propose Brain-CLIPLM, a two-stage EEG-to-text decoding framework using contrastive learning for semantic anchor extraction and a retrieval-grounded LLM with Chain-of-Thought reasoning, achieving 67.55% top-5 sentence retrieval accuracy and suggesting EEG-to-text decoding should focus on recovering compressed semantic content rather than full sentence reconstruction.
Beyond Parallel Tracking: Interactive Multi-Feature Fusion Drives Semantic Reconstruction from Non-invasive Brain Recordings
This paper introduces a multi-feature fusion framework for semantic reconstruction from non-invasive brain recordings, combining static lexical (Word2Vec) and dynamic contextual (GPT) representations via cross-attention, achieving state-of-the-art performance in brain-to-text decoding.
Continuous-Latent Predictive Modeling with Semantic Alignment for EEG-Language Foundation Models
The paper proposes BLPM, an EEG-language foundation model that uses continuous latent predictive modeling and semantic alignment to map EEG signals to text embeddings, achieving generalizable neural decoding across diverse tasks and datasets.
Margin-Regularized Structured Semantic Alignment for Brain-Language Correspondence
The paper proposes MD-SigLIP, a margin-regularized structured semantic alignment framework that directly aligns brain and text embeddings to improve brain-language decoding, achieving state-of-the-art retrieval performance.
Accurate Decoding of Natural Sentences from Non-Invasive Brain Recordings
This paper presents Brain2Qwerty v2, an AI model that accurately decodes natural sentences from non-invasive magnetoencephalography brain recordings with a 39% word error rate, demonstrating that data scaling can help bridge performance gaps with invasive methods.