Other The Verge

eBay’s bizarre cyberstalking saga ends with a $56 million settlement

eBay and former executives agreed to a $56 million settlement following a bizarre and aggressive cyberstalking campaign against a couple who ran a newsletter.

AI/ML arXiv cs.AI

xMIx: High-Performance Serving-Time Platform for Mechanistic Interpretability Apps

xMIx is introduced as a high-performance serving-native framework for deploying mechanistic interpretability applications in production AI environments with minimal overhead.

AI/ML arXiv cs.AI

An Agentic Orchestration of Atomistic Simulations

Researchers present an agent-based system within the URSA framework that automates atomistic simulations using tools like LAMMPS, reducing manual trial-and-error.

AI/ML arXiv cs.AI

HyCE-RAG: Hypergraph Chain-of-Evidence Retrieval-Augmented Generation for Explainable Multi-hop Question Answering

HyCE-RAG proposes a hypergraph-based retrieval-augmented generation framework to improve explainability and accuracy in multi-hop question answering.

AI/ML arXiv cs.AI

Differencing the Diffusion Trajectory toward Uncertain Components for Time Series Forecasting

DiffDiff is a new diffusion framework for time series forecasting that accounts for predictability asymmetry to concentrate generative effort on uncertain components.

Software Engineering Hacker News

Recursion Is Lying to You

A discussion on Hacker News regarding the misconceptions and nuances of recursion in programming.

Hardware/Chips Hacker News

I'd not buy a LG monitor

A user's negative recommendation against purchasing LG monitors based on personal experience or community sentiment.

Tech Business/VC TechCrunch

MCP startup Runlayer accuses Rippling of stealing its product idea

MCP startup Runlayer is suing Rippling for allegedly stealing its product idea after an evaluation process.

AI/ML TechCrunch

Sam Altman is ready to decelerate

Sam Altman discusses a shift in perspective toward decelerating AI development following a visceral security incident.

Other Ars Technica

New airliner sets record flying 24 hrs nonstop from Australia to France

A new airliner has set a record by flying 24 hours nonstop from Australia to France, proving the viability of Project Sunrise.

Other Ars Technica

Judge blocks first state law that would have banned prediction markets

A judge has blocked a Minnesota state law that would have banned prediction markets, allowing some types of betting to continue.

AI/ML arXiv cs.AI

TriSP: Tri-Signal Structured Pruning for Large Language Models

Introduction of TriSP, a structured pruning method for LLMs that uses a tri-signal importance metric to improve inference throughput and efficiency.

AI/ML arXiv cs.AI

ParBench: A Benchmark for Reliable Evaluation of LLM Parallel Code Translation

ParBench is a new benchmark framework for evaluating the reliability of LLM-based parallel code translation across various APIs like CUDA and OpenCL.

AI/ML arXiv cs.AI

Lexical discovery in unknown environments orchestrated by Large Language Models

The Neuro-Symbolic Lexical Discovery (NSLD) framework allows LLM agents to autonomously develop a shared vocabulary for unknown environments.

AI/ML arXiv cs.AI

Structure Over Scale: Schema-Constrained Causal Graphs for RAG

HCG-RAG introduces schema-constrained causal graphs for RAG, reducing construction costs and node count while maintaining answer quality.

Other Hacker News

The Fabled Flatbreads of Uzbekistan (2015)

A discussion or story about traditional Uzbek flatbreads.

Software Engineering Hacker News

Uv 0.12.0

Release of version 0.12.0 of uv, a fast Python package installer and resolver.

Cybersecurity Hacker News

Hulios: An eBPF-powered, transparent Tor gateway for Linux

Introduction of Hulios, a transparent Tor gateway for Linux that leverages eBPF for network routing.

Tech Business/VC The Verge

AI leaders sign statement asking the government to do something about automated AI

Employees from major AI labs call for government intervention to govern the potential automation of AI research.

AI/ML arXiv cs.AI

cMoLLM at Scale: Horizontal Scaling Laws for Mixture-of-LLMs

Proposed cMoLLM, a convolutionally gated mixture-of-LLMs that improves scaling and efficiency by treating MoE layers as dynamic convolutions.