Find a story
Search Spins
Search titles, summaries, and missing voices across published articles — press releases, announcements, and media coverage.
0 results for “neural architecture”
On the Representational Geometry of Dynamic Programs
A theoretical machine learning paper identifies geometric and algebraic reasons why neural networks fail to generalize dynamic programming tasks to longer input lengths, using tropical geometry and semiring isomorphisms to formalize structural limitations.
Aug 27, 2026
Geometry-Aware R-Structured Kolmogorov-Arnold Networks
Researchers introduced GRS-KAN, a new neural architecture that embeds differentiable R-functions into Kolmogorov-Arnold Networks to explicitly encode geometric constraints and discontinuities, improving accuracy and interpretability on benchmark regression tasks with structured boundaries.
Published Jul 3, 2026 · Analyzed Jul 6, 2026
EVOTS: Evolutionary Transformer Search for Time Series Forecasting
Researchers introduced EVOTS, an evolutionary neural architecture search framework that automatically discovers task-adaptive Transformer-like models for multivariate time-series forecasting, achieving competitive or improved MSE over fixed Transformer baselines on ETT benchmarks.
Published Jul 2, 2026 · Analyzed Jul 5, 2026