Neural Networks

A new arXiv paper presents a statistical perspective on neural architectures including feedforward networks, Deep Sets, and Transformers for amortized Bayesian inference, evaluating their accuracy, robustness, and uncert…

Neuromorphic engineering, which mimics the human brain's architecture using asynchronous processing and co-located memory and processing, promises to break the energy bottleneck of traditional von Neumann computing by en…

A 1955 proposal for the Dartmouth Summer Research Project on Artificial Intelligence, co-authored by John McCarthy, Marvin Minsky, Nathaniel Rochester, and Claude Shannon, is credited with founding the field of AI. The d…

A 2017 paper by eight Google researchers, 'Attention Is All You Need', introduced the transformer architecture, which replaced recurrence with attention mechanisms and now underpins nearly all major AI systems including …

Apple, Google, and Empirical Health have developed wearable foundation models (WFMs) — neural networks pretrained on unlabeled sensor data from devices like Apple Watch and Fitbit — that can be adapted to detect conditio…

A reverse prediction test suggests that supposedly brainlike AI models may rely on visual strategies the primate brain does not use, according to scientists. The finding indicates that the internal workings of today's vi…

A set of brain implants that bypass a spinal injury has enabled Keith Thomas, paralyzed from the chest down since 2020, to feed himself, drink from a cup, and feel his dog's fur again, according to the Guardian.

A 2026 arXiv paper introduces the concept of 'gradual cognitive externalisation,' describing how adults are incrementally offloading memory, navigation, and reasoning tasks to ambient AI systems. The authors warn this qu…

Principia Artificialis, an open mathematical research program led by user holland202, proposes that intelligence may be measurable with units, scaling laws, and phase transitions, treating machine intelligence as a physi…

A developer explains that attention sinks—tokens at position 0 that absorb excess attention weight—cause streaming LLMs to fail when evicted from the KV cache. The softmax normalization forces the model to allocate proba…

A software engineer's journey from using LLM APIs to understanding the underlying architecture reveals that models process prompts through a pipeline of tokenization, embedding, transformer layers, and iterative next-tok…

A study published in PLOS Biology on July 16, 2026, by Carta et al. used EEG recordings from normal-hearing adults in an immersive multi-talker environment to show that the brain can simultaneously encode two competing s…

Un nuevo libro de Afonso Bandeira, Amit Singer y Thomas Strohmer, 'Mathematics of Data Science', dedica un capítulo a SVD y PCA, las técnicas de álgebra lineal que permiten comprimir matrices de un millón de columnas a d…

A 135M-class depth-recurrent transformer trained on FineWeb-Edu converges to a per-token fixed point, with mean successive-output KL divergence falling from 3.9e-1 at the second loop to 8.5e-6 by the sixteenth. The resea…

Researchers have introduced CausalGraphX, a counterfactual graph neural network framework that integrates GNNs with counterfactual reasoning to provide explainable assessments of systemic risk in financial networks. The …

Researchers introduce NeuroGRIP, a retrieval-augmented graph refinement framework that uses external medical knowledge from a large-scale knowledge base derived from clinical guidelines to calibrate noisy EEG graphs for …

A new arXiv preprint proposes the Capability Convergence Hypothesis (CCH), which states that under a fixed per-token inference budget, representational convergence does not entail capability convergence; instead, capabil…

Researchers propose LIGO-PINN, a framework using learned initialization via gated layerwise optimization to overcome convergence failures in physics-informed neural networks (PINNs). In evaluations on 1D and 2D PDE domai…

Modern generative AI systems like Midjourney and Stable Diffusion recreate paintings in different artistic styles using diffusion models and attention mechanisms, according to the article. Diffusion models gradually add …

Researchers at the Feinstein Institutes for Medical Research have restored hand movement and the sense of touch to Keith Thomas, a man paralysed from the chest down, using a "double neural bypass" that combines a brain-c…