All Articles
18697 articles total
FactCheck: Feasibility-aware Long-term Action Anticipation with Multi-agent Collaboration
FactCheck is a multi-agent collaboration framework that uses an 'Observe-Plan-Verify' mechanism to improve long-term action anticipation in videos.
MatchLM2Lite: A Scalable MLLM-to-Lite Framework for Reproduced Content Identification
MatchLM2Lite is a production-grade system for reproduced content identification that distills a large multimodal model into a fast-inference student model.
Unifying Acoustic Features and Text with Multimodal LLMs for Neurodegenerative Screening
NeurMLLM is a multimodal generative framework that integrates acoustic features and text for the screening and staging of neurodegenerative diseases.
The UK's Teen Social Media Ban Is Political Theater, Not Child Safety Policy
An opinion piece arguing that the UK's proposed social media ban for teenagers is political theater rather than a substantive safety policy.
A Nipkow Disk Mechanical TV Simulator
A simulation of a Nipkow Disk, a mechanical scanning disk used in early television systems.
Anthropic’s latest feud with the Trump admin may actually help it, sales data suggests
Analysis suggests that Anthropic's tensions with the Trump administration may actually increase its business adoption and sales.
Scribby: A Multi-Level LLM Framework for Semantic Video Analysis
Scribby is a multi-level LLM framework designed for semantic video analysis, balancing macro-level comprehension with micro-level semantic chunking.
Momentum-Guided Semantic Forecasting (MoFore) for Self-Supervised Video Representation Learning
MoFore introduces a momentum-guided semantic forecasting framework for self-supervised video representation learning without action labels.
XMedFusion: A Knowledge-Guided Multimodal Perception and Reasoning Framework for Autonomous Medical Systems
XMedFusion is a modular AI framework that uses multi-agent perception and reasoning to improve radiology report generation in autonomous medical systems.
Agentomics: Economic Foundations for the Valuation, Attribution, and Pricing of AI Agents in Human-AI Workflows
Agentomics provides an economic framework for valuing, attributing, and pricing AI agents within hybrid human-AI organizational workflows using Shapley values.
An Empirical Analysis of Optimization Dynamics and Sparsity Boundaries in Large-Scale Pedestrian Attribute Recognition
An empirical study on Pedestrian Attribute Recognition (PAR) that identifies the 'Sparsity Wall' and optimizes Multi-Label Focal Loss for extreme class imbalance.
ScoutVLA: UAV-Centric Active Perception via a Dual-Expert VLA Model for Open-World Embodied Question Answering
ScoutVLA is a dual-expert Vision-Language-Action model that enables UAVs to perform active perception and evidence-seeking for outdoor embodied QA.
Double-Helix Vision (DH-V2): A Geometry-Based Visual Sampler for Bandwidth-Constrained Perception
Double-Helix Vision (DH-V2) is a geometry-based visual sampler that achieves extreme image compression for bandwidth-constrained perception without neural networks.
W.H. Auden and James Schuyler in life and literature
A discussion about the lives and literature of poets W.H. Auden and James Schuyler.
Show HN: cuTile Rust: Safe, data-race-free GPU kernels in Rust
Introduction of cuTile Rust, a library for writing safe, data-race-free GPU kernels using the Rust programming language.
Apple plans to change its Hide My Email privacy feature that could make it less effective
Apple is planning to change its Hide My Email privacy feature by moving generated addresses to a different domain, potentially reducing its effectiveness.
Cockroaches scurry around with thousands of pieces of bacterial genomes
Research reveals that cockroaches carry thousands of pieces of bacterial genomes, facilitating gene transfer across species.
Among the large new rockets Amazon was counting on, only Europe has delivered
Amazon's rocket ambitions are facing delays, with Arianespace being the only provider to have delivered on expectations so far.
Anthropic "pauses" token-based billing for its Claude Agent SDK
Anthropic has delayed the implementation of token-based billing for its Claude Agent SDK to avoid sudden cost spikes for power users.
Divide-and-Denoise: A Game-Theoretic Method for Fairly Composing Diffusion Models
Divide-and-Denoise introduces a game-theoretic method to coordinate multiple pre-trained diffusion models for fairer image composition.