Tag: interpretability

A new neuro-symbolic framework pairs context-aware particle filtering with differentiable program synthesis to infer latent cognitive states, lifting joint belief-action ...

A Plant Methods study presents a reproducible hyperspectral workflow for oat leaf chlorophyll phenotyping that reaches R² of 0.8575 with ...

A new neural architecture converts the classical Volterra series into learnable coefficient tensors, recovering known chaotic dynamics, matching transformer physics, ...

A new architecture-diverse TabNet ensemble classifies seven crop types from harmonized multi-source soil data with 90.5 percent accuracy, outperforming XGBoost ...

Concrete is one of the largest sources of human-made carbon emissions, and the race to replace it with greener alternatives ...

A new difficulty-aware Transformer framework jointly optimized by particle swarm intelligence improves sports score trend prediction while attributing forecasts to ...

Plant pathologists have long faced a stubborn diagnostic problem: a single leaf can carry several diseases at once, each with ...

A new transformer model called LitGene learns gene function by reading the scientific literature and aligning it with Gene Ontology ...

A new framework called SISTR represents word and text meanings as sparse vectors over human-interpretable topics derived from pretrained language ...

A new machine-learning framework called MB-SupConCon extends supervised contrastive learning to continuous responses, enabling more accurate and biologically interpretable prediction ...