Ciencia y Salud

ZIGAT-MAE: A Context-Conditioned Zero-Inflated Graph Attention Network with Masked Autoencoders for Spatio-Temporal Urban Crime-Risk Forecasting

Diego Saldana Ulloa

agosto 2026

Predicting urban crime struggles with sparse data, spatial dependencies, and variable log reliability. We propose ZIGAT-MAE, a graph attention network on H3 hexagonal grids. It fuses dynamic, static, and environmental data via dual-layer GAT and FiLM conditioning, outputting Zero-Inflated Poisson (ZIP) parameters through dual heads to handle differing data streams, optimized by a sparsity-conscious loss. The model uses multi-city masked autoencoder (MAE) pretraining to reconstruct static features and "ghost-imputation" to drop untrustworthy data. Evaluated across 9 Mexican metros, ZIGAT-MAE achieves a 0.942 ROC-AUC, 194x baseline average precision, and excellent calibration errors (under 10^-3). Targeting the highest-risk 5% of areas captures 72% of incidents, beating historical baselines. Ablations confirm MAE pretraining and ZIP fine-tuning are synergistic. Extreme crime clustering (Gini 0.96; top 5% of areas see 89% of crime) validates the architecture for strategic resource deployment.

Reading Without a Reader: Large Language Models Collapse Reading and Writing into a Single Entangled Code

Diego Saldana Ulloa

julio 2026

In the literate human brain, reading and writing doubly dissociate: a ventral decoding route (pure alexia) and a fronto-parietal encoding route (pure agraphia), sharing a partial orthographic core. A decoder-only large language model (LLM) drives both from one autoregressive path optimized on text (a \emph{cultural} invention, not an evolved instinct). We ask how entangled it is, comparing an input-side ``reading code'' WE with an output-side ``writing code'' WU via an index E∈[0,1] (CKA, Procrustes residual, mutual k-NN) calibrated against an independent-init floor and tied ceiling. On GPT-2, OPT and Pythia (14M--1.4B), untied models hold one \emph{coupled but sub-ceiling} code (E=0.23--0.35, far above floor) on a non-monotonic couple-then-differentiate trajectory, WU drifting ∼3.2× farther than WE in every decile. Equally informative is a negative: the matching behavioural test, that comprehension and production fail together rather than dissociate, cannot be run. For minimal pairs the alexia analogue is em

Tócate para que no te toque

Sandra Edith Nava Munoz

octubre 2025

El Cáncer de Mama es el tipo de cáncer con mayor incidencia en mujeres. El área médica, para una detección oportuna de la enfermedad sugiere lo siguiente: i) a partir de los 20 años realizar una autoexploración física, ii) a partir de los 30 años...

Tecnología en Apoyo al Cuidado del Adulto Mayor

Sandra Edith Nava Munoz

septiembre 2025

El cuidado del adulto mayor es un proceso que presenta desafíos. Las ciencias de la computación pueden intervenir de diferentes formas, creando entornos adecuados, diseñando sistemas para las necesidades del adulto mayor.

The evolution of exceptional diversity in parental care and fertilization modes in ray-finned fishes

Oscar Garcia Miranda

septiembre 2025

Ray-finned fishes show diverse parental care. From 7,600 species, we found transitions linked to fertilization: oviductmouth favor female care, brood pouch male care. Internal fertilization without care is unstable, revealing aquatic-driven flexibility.

El porqué de la coloración animal

Oscar Garcia Miranda

septiembre 2025

La coloración animal, estudiada por Darwin, Wallace y múltiples disciplinas, sigue siendo un enigma fascinante: un espectáculo natural aún sin respuesta definitiva.