Data & Analytics

Zero-shot imagined speech decoding from MEG via imagined-to-listened cross-condition mapping. Trains models to map imagined MEG responses to listened responses, then decodes using listened-only decoder. Three-stage pipeline: (1) mapping imagined→listened MEG, (2) train contrastive word decoder on listened MEG with multi-embedding evaluation, (3) decode imagined speech via mapping pipeline on held-out subjects. Use when: imagined speech decoding, MEG BCI, cross-condition neural mapping, zero-s...

YANA: Bridging the Neuromorphic Simulation-to-Hardware Gap. Framework for seamless translation of SNN algorithms from simulation to neuromorphic hardware deployment. Activation: YANA, simulation-to-hardware, neuromorphic deployment, SNN hardware gap.

Skill for applying Cluster-based Sequential Feature Selection (CSFS) to improve feature selection in wind and solar power prediction tasks. Use when working with renewable energy prediction datasets that have many environmental variables and need efficient, model-agnostic feature selection.

Wavelet-Enhanced Mixture-of-Experts (WaveMoE) foundation model for time series forecasting. Use when building time series prediction models, incorporating frequency-domain information, or designing MoE architectures for temporal data.

Multi-Stage Warm-Start (MSWS) deep learning framework for Unit Commitment optimization. Combines neural network warm-starting with MILP constraints to accelerate power grid scheduling. Use for unit commitment, power system optimization, energy scheduling, and MILP warm-starting.

Beyond-symmetry structural design patterns for variational quantum machine learning. Use when designing VQML ansatze that go beyond symmetry constraints, selecting parametrizations that balance expressivity and trainability within symmetry-preserving subspaces, or analyzing structural choices in quantum neural network architectures. Covers equivariant VQA design, symmetry-breaking regularization, and structural ansatz selection criteria.

--- name: visual-cortex-diffusion-model description: "Skill for understanding and applying the mechanistic model of inference in visual cortex equivalent to a minimal diffusion model, linking sparse coding with recurrent dynamics and horizontal connections in V1. Based on arXiv:2607.15693." activation: visual cortex diffusion model, sparse coding inference, recurrent diffusion model

Skill for understanding and applying the mechanistic model of perceptual inference in visual cortex equivalent to a minimal diffusion model (arXiv:2607.15693). Enables extraction of principles linking sparse coding, recurrent dynamics, and diffusion model training for neuroscience-inspired machine learning.

Treatment-Conditioned Diffusion framework for forecasting neurodegenerative disease progression via high-fidelity brain state prediction. Conditions generative process on DaTscan images and levodopa equivalent daily dose. Activation: neurodegenerative, disease progression, Parkinson, diffusion, longitudinal neuroimaging, DaTscan, treatment-conditioned.

Information geometry framework for analyzing non-Euclidean structure of visual space — modeling perceptual geometry using Riemannian manifolds, Fisher information, and Finsler geometry. Activation: visual space, non-Euclidean, information geometry, Riemannian manifold, perceptual geometry, Fisher information, visual perception, psychophysics.

GNN-based visual category decoding from fMRI functional brain networks — signed graph neural network with sparse edge masks and class-specific saliency for decoding sports, food, and vehicles from 7T fMRI data. Activation: fMRI decoding, visual category, GNN, brain network, graph neural network, saliency, functional connectivity.

GFlowState visual analytics system for illuminating Generative Flow Network training. Provides interactive visualizations of training states, flow distributions, and mode coverage for diagnosing mode collapse and reward hacking. Activation: GFlowState, GFlowNet visualization, generative flow networks, training visualization.

Generative model for bipartite gene-sharing networks explaining evolutionary patterns in viruses and mobile genetic elements. Captures scale-free gene degree and exponential genome degree distributions via horizontal gene transfer, gene capture, and loss processes. Activation: gene-sharing network, bipartite network evolution, viral genome evolution, pangenome modeling, horizontal gene transfer.

Eccentricity-Constrained CNN Training methodology for adaptive visual information coding around the visual field using egocentric data

Eccentricity-constrained CNN training on egocentric video reveals adaptive, task-aligned information coding across the visual field, with fovea-preferred models advantaged for face and object tasks and periphery-preferred models favored in scene-selective cortex.

Skill for AI agent capabilities

Derived from arXiv:2607.17281 - AIGB-R1: Self-Evolving Generative Auto-Bidding via Hierarchical Planner-Executor Optimization

Variational framework for statistical inference on cyclic interactions in directed networks. Directed interactions as edge flows on simplicial complex evolved under energy-minimizing dynamics, yielding low-dimensional cycle space for recurrent organization. Activation: cyclic interaction, harmonic flow, cycle space, simplicial complex, recurrent network, directed graph cycles.

Variational Quantum Algorithms methodology covering CVQE (Cascaded Variational Quantum Eigensolver), certified QNN training via QIBP, and resource-efficient quantum optimization. Use when designing variational quantum circuits, optimizing NISQ-era algorithms, implementing certified quantum machine learning, or applying quantum algorithms to combinatorial optimization problems. Covers VQE variants, quantum interval bound propagation, compact binary encoding for quantum optimization, and divide...

Systematic RSA comparison showing untrained CNNs match backpropagation-trained CNNs at V1 visual cortex. Untrained random-weights CNN (rho=0.076) exceeds backprop (rho=0.034) at V1/V2 (p<0.001). STDP achieves highest V1 alignment among trained rules (rho=0.064). Four learning rules (BP, FA, PC, STDP) compared against human fMRI from THINGS-fMRI dataset (720 stimuli, 3 subjects).

Systematic RSA comparison showing untrained CNNs match backpropagation at V1 alignment with human fMRI. Evaluates BP, FA, PC, and STDP learning rules against THINGS-fMRI dataset using 720 stimuli across 3 subjects. Use when studying brain-model alignment, comparing learning rules, or analyzing visual cortex representations via Representational Similarity Analysis.

Systematic RSA comparison showing untrained CNNs match backpropagation-trained CNNs at V1 visual cortex. Trigger words: untrained CNN, backpropagation, RSA, V1, representational similarity

Systematic RSA comparison showing untrained CNNs match backpropagation-trained networks at V1 visual cortex, revealing architecture's dominant role over learning rules in neural alignment. Activation triggers: untrained cnn, backpropagation, v1, rsa, representational similarity, learning rules, architecture-driven.

Systematic RSA comparison showing untrained CNNs match backpropagation-trained CNNs in V1 visual cortex alignment. Large-scale fMRI analysis reveals that random feature detectors can capture V1 representational structure. Keywords: untrained CNN, V1 cortex, backpropagation, RSA, representational similarity, visual cortex, fMRI.