Neural Networks

Researchers evaluated UAV detection and tracking under synthetic fog, finding that fog degrades performance primarily through missed detections. Fog-inclusive training improved robustness most consistently, while test-ti…

Researchers introduced a rendering-aware Bayesian 3D Gaussian splatting framework that provides native uncertainty estimates and adaptive complexity control, outperforming standard ensembles in active view selection task…

Researchers developed a reduced NEXI protocol for quantifying human gray matter microstructure on the Connectome 2.0 scanner, using an Explainable AI framework to cut scan time from 27 to 14 minutes. Validated in seven p…

Researchers introduced a Tri-Branch Modular Fusion Neural Network that combines 3D geometry, SMILES embeddings, and physicochemical descriptors for molecular property prediction. The model achieved a validation MAE of 0.…

Researchers proposed a controllability-observability framework for compressing deep neural networks by viewing them as nonlinear dynamical systems. The method achieved up to 83% parameter compression on CIFAR-10 with neg…

Researchers propose CcGAN-AVAR, an imbalance-robust extension of Continuous conditional GAN that uses adaptive vicinal learning and auxiliary regularization to improve generation quality and label consistency. The method…

Researchers show that incorrect L0 hyperparameter settings in Sparse Autoencoders (SAEs) cause feature mixing and degenerate solutions, preventing proper disentanglement of LLM internal features. They propose a proxy met…

Researchers derived smooth Monte-Carlo estimators for Minkowski functionals of neural fields, achieving 1-3% accuracy in 2D and 3D topology estimation at 250x speedup over persistent homology. However, in 3D neural-SDF f…

Researchers at arXiv introduced HARVEY, a defense against neural backdoor attacks that learns a reference model for poisonous samples instead of benign ones. By identifying poisonous samples more accurately than prior me…

Researchers introduced Algorithmic Core Extraction (ACE) to isolate compact subspaces, or algorithmic cores, that are necessary and sufficient for a task and recur across independently trained transformers. Across synthe…

Researchers propose IMR, a novel method for multi-agent trajectory prediction that combines mode-world weighted regression loss and an iterative decoder to improve mode diversity and prediction accuracy. The method ranks…

A developer argues that the real bottleneck in AI development is not GPU shortages but a fundamental issue identified in 2007: the lack of efficient algorithms for training large neural networks. The post highlights that…

Polysemanticity in neural networks arises from superposition, where a single neuron activates for multiple distinct inputs due to insufficient neurons. In language models, this enables efficient representation of compoun…

Arno Schwarz released 'Descent', an EGA-style point-and-click adventure game set inside a neural network's hidden layers. Players navigate through recognition, judgment, and recursion using a verb-driven parser, with cho…

A new analysis explains how predictive processing shapes the content of visual hallucinations in Charles Bonnet Syndrome, suggesting that learned expectations sculpt bursts of neural activity into persistent hallucinatio…

Researchers at Goodfire, Harvard, Stanford, and Brown introduced Block-Sparse Featurizers (BSF), a method that decomposes neural network activations into multidimensional subspaces instead of single directions. Applied t…

Ilya Sutskever's curated list of 30 foundational machine learning papers, originally shared with game developer John Carmack in 2019, has been organized into a structured roadmap at 30papers.com. The curriculum spans fro…

Intelligence can lead to poor decisions as smart people become adept at justifying irrational choices through motivated reasoning. Psychologists find that higher cognitive ability often correlates with stronger confirmat…

Some AI researchers argue that scaling large language models may not be sufficient to achieve human-level intelligence, pointing to alternative approaches such as world models, pure reinforcement learning, or neurosymbol…

A bioRxiv preprint covered by Neuroscience News and PsyPost on July 7, 2026, reports that researchers recorded 79 infants with simultaneous EEG and DeepLabCut pose tracking, finding neural sensitivity to structured music…