Research map / Large language models / Language, safety and interpretability
Towards Sentence-Level Brain Decoding with Distributed Representations
Jingyuan Sun, Shaonan Wang, Jiajun Zhang, Chengqing Zong
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| structured supervised model InferSent instead of Avg | sentence representation · same-passage matching, Ridge | matching accuracy: 0.73 -> 0.83 | Improved |
| InferSent versus Cat | sentence representation · same-passage matching, MLP | matching accuracy: 0.66 -> 0.79 | Improved |
| unstructured pooling versus structured models on coarse topic task | Avg/Cat · different topic, Ridge | matching accuracy: 0.92 vs 0.94 | No clear effect |
| concatenate average and max pooling instead of single pooling | Cat · same passage, Ridge | matching accuracy: 0.65 -> 0.74 | Improved |
| SIF weighted averaging instead of naive averaging | SIF · similarity-based decoding | matching accuracy | No clear effect |
| Skip-thought structured model instead of Cat on same-passage task | Skip-thought · same passage, Ridge | matching accuracy: 0.74 -> 0.75 | Improved |
| MLP regression instead of Ridge for Cat | Cat · same passage | matching accuracy: 0.74 -> 0.66 | Hurt |
| longer sentences (ablation) | sentence representations · per-sentence decoding | decoding accuracy | Improved |
| rearranging word order while preserving meaning | structured models · brain decoding | decoding accuracy | No clear effect |
| InferSent versus GenSen on rank evaluation | InferSent · rank accuracy | ranking score | Improved |
Closest papers by meaning
- Do Large Language Models Think like the Brain? Sentence-Level Evidences from Layer-Wise Embeddings and fMRI (AAAI 2026 · similarity 0.90)
- Inducing brain-relevant bias in natural language processing models (NeurIPS 2019 · similarity 0.90)
- Disentangling syntax and semantics in the brain with deep networks (ICML 2021 · similarity 0.90)
- Far from the Shallow: Brain-Predictive Reasoning Embedding through Residual Disentanglement (NeurIPS 2025 · similarity 0.89)
- fMRI predictors based on language models of increasing complexity recover brain left lateralization (NeurIPS 2024 · similarity 0.87)
- Improving Semantic Understanding in Speech Language Models via Brain-tuning (ICLR 2025 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
