Researchers introduced a Gibbs posterior sampler for solving inverse problems using a diffusion-based prior model. The method addresses ill-posed linear inverse problems with Bayesian regularization, …
Researchers proposed a low-overhead error-correction technique for quantum convolutional neural networks (QCNNs) using bivariate bicycle (BB) codes. Simulations showed that an unprotected 4-qubit QCNN…
Researchers proposed GAIDE, a framework to help K-12 teachers create AI-powered learning tools using vibe coding. Validated through an eight-week workshop with three teachers and four mentors, the fra…
Researchers introduced Reflexion, an AI-powered prototype for structured emotional self-reflection that integrates emotion detection, layered prompting, and metaphorical storytelling. The system guide…
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. …
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 accu…
A new arXiv paper demonstrates that standard importance weighting fails to correct for selection bias in machine learning when selection depends on unobservables correlated with outcomes. The authors …
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 accura…
Researchers propose retrieval-augmented generation and constrained decoding to improve LLM-generated web API invocation code. RAG reduces hallucinations but can introduce unnecessary parameters, while…
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…
Researchers proposed a method to estimate observation parameters and noise in inverse problems using a diffusion-based prior model within a Bayesian framework. The approach enables optimal estimation …
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 transfo…
Apple is in early-stage talks with Chinese memory chip maker CXMT to secure DRAM supply for its devices, marking a sharp turnaround for the state-backed company that is central to Beijing's AI supply …
CNBC's Jim Cramer said Samsung's earnings sparked a sell-off in AI hardware stocks, while many tech names that have lagged this year rebounded, potentially signaling a shift in AI leadership.…
Developer Ashutosh Rath released Atrophy, a command-line tool that treats coding skills like Elo chess ratings and offers drills in five areas to combat skill decay from over-reliance on AI coding ass…
Security researchers at Noma Labs discovered a prompt injection vulnerability in GitHub's Agentic Workflows, dubbed GitLost, that allows attackers to trick AI agents into leaking private repository da…
Samsung Electronics reported earnings that fell short of high AI-driven expectations, triggering a sell-off in chip stocks after the company's shares had surged 145%.…
Meta released its first image generation model, Muse Spark Image, since CEO Mark Zuckerberg's AI overhaul. The model will be integrated into Meta's AI chatbot and Instagram photo app, marking a key st…
Meta launched Muse Image, its first AI model for image generation, aiming to attract advertisers and creators to its platform.…
Samsung's record earnings report triggered a sell-off in the US chip sector on Tuesday, as investors worried about overspending on AI infrastructure and lofty expectations. Shares of Intel, Marvell, a…