Geometric stability of neural population codes, quantified by split-half RDM Spearman correlation, is behaviorally relevant, regionally variable opposite to temporal stability, and supported by recurrent excitatory coupling in attractor networks.
hub
Representational similarity analysis – connecting the branches of systems neuroscience , issn =
24 Pith papers cite this work, alongside 1,292 external citations. Polarity classification is still indexing.
24
Pith papers citing it
1,292
external citations · Crossref
hub tools
citation-role summary
background 1
baseline 1
citation-polarity summary
representative citing papers
Introduces a Q-sort protocol using human reference factors to quantify LLM value-structure alignment via Procrustes similarity and RSA correlations, revealing cross-family heterogeneity and localized misalignments.
Tensor similarity is a symmetry-invariant metric that measures functional equivalence between tensor-based networks using a recursive algorithm for cross-layer mechanisms.
A mixture model with adaptive KDE and per-image cross-validation raises estimated human fixation consistency by 5-15% median log-likelihood and up to 2 AUC points over fixed-bandwidth Gaussian baselines.
SRA reframes CTKD by aligning attention-weighted span centers of mass in a multi-particle system model with geometric regularization and span logit distillation, claiming consistent outperformance over baselines.
Meditation is proposed to increase functional signal-to-noise ratio (f-SNR) in the brain via signal enhancement, noise decluttering, reduced self-referential filtering, and shifts to critical neural regimes, unifying diverse findings and explaining benefits across psychopathologies.
A PC-based decomposition of FVE into low- and high-dimensional components reduces bias when applying GWASH or LMM-REML to strongly correlated high-dimensional predictors.
Sparse autoencoders applied to GPT-2 and Llama models recover semantic features accounting for 94% of peak brain encoding performance and map onto distinct cortical semantic regions across three languages.
Transformer represents but does not causally transmit staged algorithmic intermediates for base-digit extraction, diverging from probe predictions.
Subject-specific fMRI embeddings learned unsupervised from the Natural Scenes Dataset can be aligned across individuals via orthogonal rotations, supporting a shared neural geometry in visual cortex.
The authors propose target-space recovery profiles to diagnose which reproducible dimensions of fMRI brain responses are captured by model predictions, showing that accuracy alone can mask alignment mismatches in visual cortex.
Heterogeneous visual agents form shared symbols via decentralized Metropolis-Hastings captioning, where encoder similarity shapes the content and symmetry of the resulting language.
Decoding alignment metrics can remain high and unchanged even when encoding manifold topology is causally altered, so they do not imply similar function or computation across neural populations.
LLMs exhibit Bayesian-like hypothesis updating with strong-sampling bias and an evaluation-generation gap but generalize poorly outside observed data.
Alignment pattern analysis reveals that models aligned to individual brain ROIs do not reproduce the stable cross-region alignment profiles observed across human subjects.
Language models encode concept hierarchies as linear transformations that are domain-specific yet structurally similar across domains.
A decentralized collective world model integrates predictive coding with bidirectional communication to achieve simultaneous symbol emergence and coordination, outperforming non-communicative baselines in a two-agent trajectory task under divergent perceptions.
The paper introduces a time-resolved neural encoder combining Whisper embeddings with recurrent temporal modeling and soft attention to predict ECoG responses, finding strongest alignment in intermediate layers and anatomically coherent phoneme organization in electrodes.
MTA is a distillation method that aligns teacher-student LLM representations along their transformation trajectories using layer-adaptive granularities and dynamic structural plus hidden representation alignment losses.
No correlation exists between CNNs' Brain-Score alignment with the visual system and the perceptual content of their Gram-matrix texture representations.
RSA on 7T fMRI during natural scene viewing identifies ventromedial and lateral occipitotemporal representational routes for scene context versus animate content, with differential alignment to vision and language models.
TALAS is a knowledge distillation method that selectively aligns upper student layers to teacher sentence embeddings, propagates knowledge top-down via relational constraints in lower layers, and uses ASAM to seek flatter minima.
Probing classifiers are a common but limited method for analyzing linguistic knowledge in neural NLP models, and this review outlines their promises, methodological shortcomings, and recent advances.
citing papers explorer
-
Geometric Stability of Neural Population Codes: Regional Variation, Behavioral Relevance, and Circuit Dependence
Geometric stability of neural population codes, quantified by split-half RDM Spearman correlation, is behaviorally relevant, regionally variable opposite to temporal stability, and supported by recurrent excitatory coupling in attractor networks.
-
Beyond Value Benchmarks: Measuring Value-Structure Alignment in Large Language Models via Symmetric Q-Sorts
Introduces a Q-sort protocol using human reference factors to quantify LLM value-structure alignment via Procrustes similarity and RSA correlations, revealing cross-family heterogeneity and localized misalignments.
-
When Are Two Networks the Same? Tensor Similarity for Mechanistic Interpretability
Tensor similarity is a symmetry-invariant metric that measures functional equivalence between tensor-based networks using a recursive algorithm for cross-layer mechanisms.
-
Raising the Ceiling: Better Empirical Fixation Densities for Saliency Benchmarking
A mixture model with adaptive KDE and per-image cross-validation raises estimated human fixation consistency by 5-15% median log-likelihood and up to 2 AUC points over fixed-bandwidth Gaussian baselines.
-
SRA: Span Representation Alignment for Large Language Model Distillation
SRA reframes CTKD by aligning attention-weighted span centers of mass in a multi-particle system model with geometric regularization and span logit distillation, claiming consistent outperformance over baselines.
-
Clear Mind: Meditation and the Brain's Signal-to-Noise Ratio
Meditation is proposed to increase functional signal-to-noise ratio (f-SNR) in the brain via signal enhancement, noise decluttering, reduced self-referential filtering, and shifts to critical neural regimes, unifying diverse findings and explaining benefits across psychopathologies.
-
Principal Components Decomposition of Fraction of Variance Explained in High Dimensional Linear Models with Strong Correlation
A PC-based decomposition of FVE into low- and high-dimensional components reduces bias when applying GWASH or LMM-REML to strongly correlated high-dimensional predictors.
-
Sparse Autoencoders Map Brain-LLM Alignment onto Cortical Semantic Topography
Sparse autoencoders applied to GPT-2 and Llama models recover semantic features accounting for 94% of peak brain encoding performance and map onto distinct cortical semantic regions across three languages.
-
Represented Is Not Computed: A Causal Test of Candidate Algorithmic Intermediates in a Transformer
Transformer represents but does not causally transmit staged algorithmic intermediates for base-digit extraction, diverging from probe predictions.
-
Platonic Representations in the Human Brain: Unsupervised Recovery of Universal Geometry
Subject-specific fMRI embeddings learned unsupervised from the Natural Scenes Dataset can be aligned across individuals via orthogonal rotations, supporting a shared neural geometry in visual cortex.
-
Beyond Prediction Accuracy: Target-Space Recovery Profiles for Evaluating Model-Brain Alignment
The authors propose target-space recovery profiles to diagnose which reproducible dimensions of fMRI brain responses are captured by model predictions, showing that accuracy alone can mask alignment mismatches in visual cortex.
-
Emergent Communication between Heterogeneous Visual Agents through Decentralized Learning
Heterogeneous visual agents form shared symbols via decentralized Metropolis-Hastings captioning, where encoder similarity shapes the content and symmetry of the resulting language.
-
Decoding Alignment without Encoding Alignment: A critique of similarity analysis in neuroscience
Decoding alignment metrics can remain high and unchanged even when encoding manifold topology is causally altered, so they do not imply similar function or computation across neural populations.
-
Hypothesis generation and updating in large language models
LLMs exhibit Bayesian-like hypothesis updating with strong-sampling bias and an evaluation-generation gap but generalize poorly outside observed data.
-
Only Brains Align with Brains: Cross-Region Alignment Patterns Expose Limits of Normative Models
Alignment pattern analysis reveals that models aligned to individual brain ROIs do not reproduce the stable cross-region alignment profiles observed across human subjects.
-
Linear Representations of Hierarchical Concepts in Language Models
Language models encode concept hierarchies as linear transformations that are domain-specific yet structurally similar across domains.
-
Decentralized Collective World Model for Emergent Communication and Coordination
A decentralized collective world model integrates predictive coding with bidirectional communication to achieve simultaneous symbol emergence and coordination, outperforming non-communicative baselines in a two-agent trajectory task under divergent perceptions.
-
Mapping Whisper Representations to Human ECoG Responses with Interpretable Time-Resolved Neural Encoding
The paper introduces a time-resolved neural encoder combining Whisper embeddings with recurrent temporal modeling and soft attention to predict ECoG responses, finding strongest alignment in intermediate layers and anatomically coherent phoneme organization in electrodes.
-
MTA: Multi-Granular Trajectory Alignment for Large Language Model Distillation
MTA is a distillation method that aligns teacher-student LLM representations along their transformation trajectories using layer-adaptive granularities and dynamic structural plus hidden representation alignment losses.
-
Perceptual misalignment of texture representations in convolutional neural networks
No correlation exists between CNNs' Brain-Score alignment with the visual system and the perceptual content of their Gram-matrix texture representations.
-
Shared representations in brains and models reveal a two-route cortical organization during scene perception
RSA on 7T fMRI during natural scene viewing identifies ventromedial and lateral occipitotemporal representational routes for scene context versus animate content, with differential alignment to vision and language models.
-
TALAS: Teacher-Anchored Layer Alignment with Adaptive Sharpness-Aware Minimization for Embedding Distillation
TALAS is a knowledge distillation method that selectively aligns upper student layers to teacher sentence embeddings, propagates knowledge top-down via relational constraints in lower layers, and uses ASAM to seek flatter minima.
-
Probing Classifiers: Promises, Shortcomings, and Advances
Probing classifiers are a common but limited method for analyzing linguistic knowledge in neural NLP models, and this review outlines their promises, methodological shortcomings, and recent advances.
- Semantic knowledge guides innovation and drives cultural evolution