Software Engineering arXiv cs.AI

DragonCrawl: A Generative, Intent-Based Framework for Scalable Mobile End-to-End Testing

Presents DragonCrawl, an AI-driven mobile E2E testing framework that uses generative intent-based reasoning to reduce test maintenance.

AI/ML arXiv cs.AI

SCMA: Structure-Conditioned and Metal-Aware Flow Matching for CT Metal Artifact Reduction

Proposes SCMA, a flow matching framework designed to reduce metal artifacts in CT scans while preserving anatomical structure.

AI/ML arXiv cs.AI

WaiT for the Signal: Simple Frequency-Aware Flow-Matching

Introduces WaiT, a Wavelet-aware image Transformer that improves high-resolution image and video generation efficiency and quality.

AI/ML arXiv cs.AI

Stratified Negation in RDF Rules: A Correct Approach (Extended Version)

Discusses a new approach called chain stratification to handle negation in RDF rule languages like N3 and SHACL.

AI/ML arXiv cs.AI

Benchmarks Are Not Monolithic: Sample-Level Auditing and Orchestration for LLM Evaluation

A framework for sample-level auditing of LLM benchmarks to move beyond monolithic accuracy scores toward targeted capability evaluation.

AI/ML arXiv cs.AI

Rolling With Resistance: Preference-Optimized LLM Counselors Can Trade Goal Persistence for Relational Attunement in Motivational Interviewing

Analyzes the trade-off between goal persistence and relational attunement in LLM counselors using preference optimization.

AI/ML arXiv cs.AI

Hypergradient-based Bilevel Reinforcement Learning with Improved Sample Complexity

Proposes a Hessian-free hypergradient-based bilevel RL algorithm with improved sample complexity for meta-learning and RL-HF.

Software Engineering Hacker News

Bonsai: Janestreet's UI Library

Jane Street has released Bonsai, a UI library designed for their specific internal needs.

Other Hacker News

The Shape of Things to Come

A discussion on future technological trends and the trajectory of software and hardware evolution.

Software Engineering Hacker News

How do you rewrite C/C++ projects to Rust?

A community discussion exploring strategies and best practices for migrating legacy C/C++ codebases to Rust.

AI/ML Hacker News

The AI Productivity Gap

An exploration of the productivity gap created by AI tools and how it affects developer efficiency and output.

Tech Business/VC Hacker News

Situational Awareness and the Impending Stock Market Volatility

An analysis of current market conditions and predicted stock market volatility influenced by systemic AI-driven shifts.

Software Engineering Hacker News

Show HN: I created a project management system

A developer showcases a custom-built project management system designed for streamlined task tracking.

AI/ML arXiv cs.AI

Reflected UAS: Corrected Deterministic Stability and Direct CTMC Drift Calculation

Research on Reflected UAS routing for multi-server queues, providing a direct Foster-Lyapunov drift inequality for CTMC stability.

AI/ML arXiv cs.AI

Code Is the Body: Agent-Owned Software Bodies for Recursive Evolution and Descent

Introduction of OurArk, an architecture for persistent personal AI agents with agent-owned software bodies for recursive evolution.

AI/ML arXiv cs.AI

SEDR-Seq2P: A Lightweight Dilated Residual Sequence-to-Point Network for Multi-Task Industrial NILM

Proposed SEDR-Seq2P, a lightweight dilated residual sequence-to-point network for industrial Non-Intrusive Load Monitoring (NILM).

AI/ML arXiv cs.AI

Predicting Steel Fatigue Life from Micrographs Using Physics-Informed Deep Learning

A physics-informed deep learning framework using computer vision to predict the fatigue life of alloy steels from micrographs.

AI/ML Hacker News

Emulating ALiBi with Rope

A discussion on Hacker News regarding the emulation of ALiBi (Attention with Linear Biases) using RoPE (Rotary Positional Embeddings) in LLMs.

AI/ML arXiv cs.AI

Unanticipated Effects of Generative AI on Expertise Pathways and Performance Perception in System Administration

Research exploring how generative AI affects the development of technical expertise and performance perceptions in system administration.

AI/ML arXiv cs.AI

HenTwin: A Multimodal Digital Twin Framework for Longitudinal Biological State Monitoring in Laying Hens

Introduction of HenTwin, a multimodal digital twin framework using a five-layer IoT architecture for biological state monitoring in laying hens.