All Articles
18657 articles total
Show HN: StarScope – Free astronomy dashboard for observers outside the US/UK
StarScope is a free astronomy dashboard designed for observers located outside the US and UK.
FTC lawsuit reveals how subscription scam networks evade app store enforcement
An FTC lawsuit exposes how subscription scam networks use shell companies to bypass app store enforcement and continue operating.
Trust-Aware Multi-Agent Traceability: Confidence-Calibrated Knowledge Graphs for Consistent Software Artifact Management
Researchers propose a trust-aware coordination framework using calibrated confidence scores and knowledge graphs to improve multi-agent AI software engineering pipelines.
Rift: A Conflict Signature for Deception in Language Models
The 'Rift' research identifies a specific internal 'conflict signature' in LLMs that distinguishes deceptive responses from honest errors or hallucinations.
Physics-Informed Attention Mechanism and Generalization Capability of Deep Learning-Based Grain Growth Evolution Prediction
A study demonstrates that physics-informed boundary-masked attention mechanisms improve the OOD generalization of ML models predicting grain growth evolution.
MLLP-VRAIN UPV system for the IWSLT 2026 Simultaneous Speech Translation task
The MLLP-VRAIN group presents a cascaded Simultaneous Speech Translation system using Parakeet and Qwen 3.5 models with adaptive policies and RAG.
Pulling The REINS: Training-Free Safety Alignment of Video Diffusion Models via Representation Steering
REINS is a training-free safety alignment method for video diffusion models that steers internal representations to prevent the generation of unsafe content.
ARVO: Atlas of Reproducible Vulnerabilities for Open-Source Software
ARVO is a new large-scale dataset of over 6,100 reproducible real-world vulnerabilities in open-source software, designed to aid security research.
From Democracies to Autocracies: How AI Systems Enable Authoritarianism by Design
A qualitative study maps how AI systems are designed and deployed in various political regimes to enable authoritarianism through data centralization and governance gaps.
Transformer-Based Warm-Starting for Feasible and Optimal Terminal Approach to Tumbling Objects with Space Manipulators
A transformer-based warm-starting approach is used to reduce the computational runtime and improve the robustness of trajectory generation for space manipulators.
The hacker sent by Anthropic to calm the government's nerves about AI safety
Discussion regarding an Anthropic employee tasked with addressing government concerns about AI safety.
World leaders want American AI. They just don’t want America to be able to turn it off.
World leaders express concern over U.S. control of AI infrastructure, fearing sudden access termination.
Epic wants to let you bring your Fortnite skins to other games
Epic Games plans to introduce interoperable Fortnite skins in Unreal Engine 6, allowing assets to move between different games.
AI coding agents can autonomously direct robot training
NVIDIA is utilizing AI coding agents to autonomously manage and direct the training process for robots.
Anthropic ships major Claude Design overhaul with design system imports, code round-trips, and a fix for its token-burning problem
Anthropic updates Claude Design with design system imports and bidirectional integration with Claude Code to streamline the design-to-engineering handoff.
Agentic Discovery of Non-Canonical Antimicrobial Peptides with AMPGAN v3
Researchers introduce AMPGAN v3 for generating antimicrobial peptides and PepCraft, a multi-agent framework for end-to-end peptide discovery.
PromptMN: Pseudo Prompting Language
PromptMN is proposed as a pseudo-prompting domain-specific language to reduce ambiguity in human-AI interactions through typed directives.
Statistical Foundations of LLM-based A/B Testing: A Surrogacy Framework for Human Causal Inference
A new statistical framework investigates the validity of using LLMs as surrogates for human participants in A/B testing.
Cluster-Aware Dual-Level Test Specification Generation for Large-Scale Automotive Software Requirements
A 'Cluster-then-Summarize' pipeline is developed to automate test specification generation for large-scale automotive software requirements.
PowerOPD: Stabilizing On-Policy Distillation with Bounded Power Transformation
PowerOPD is introduced to stabilize on-policy distillation in LLMs using a bounded power transformation to reduce gradient variance.