AI News

Lead Sparker (www.producthunt.com)

Veille quotidienne — 18/09/2026

Staff AI (www.producthunt.com)

Jemalloc 5.4.0 (github.com)

The Scourge of x86 Emulation (fex-emu.com)

A Unified Evaluation Framework for Trustworthy Large Language Models, Agentic AI, and Multimodal Systems (arxiv.org)

The First-Order Oracle Complexity of Lipschitz Convex Optimization in Nondual Settings (arxiv.org)

Beyond PINNs: A Unified Gauss--Newton and Petrov--Galerkin Framework for Neural and Hybrid PDE Solvers (arxiv.org)

NS3Learn: Transferring 5G NR Mode-2 Reception Realism from ns-3 to the Veins/SUMO Stack for Connected-Vehicle Safety Assessment (arxiv.org)

Relational Attention for Data-Efficient Language Modeling (arxiv.org)

Noise-Robust Quantum State Characterization for Remote State Preparation with Deep Learning (arxiv.org)

Behavioral Coherence: A Method for Sensitive-Domain LLM Evaluation (arxiv.org)

Truncated automatic sparse differentiation for machine learning interatomic potentials (arxiv.org)

Coding Agents with an Obstacle-Aware Harness for Safe Robot Manipulation (arxiv.org)

Workspace Models: Lightweight Robotic Memory via Saliency-Driven Supervision (arxiv.org)

Correlation-Free Transition Path Sampling through Shooting Point Generation Guided by Committor Learning (arxiv.org)

Cross-Architecture Foundation-Model Distillation for Edge Flood Segmentation (arxiv.org)

Detecting Deceptive Recruitment: A Signal-theoretic Machine Learning Framework for Early Identification of Labour Exploitation (arxiv.org)

Paint-Anything: Unified Any-Color Control for Image Generation and Editing (arxiv.org)

ERCPMP-Gx: Endoscopic Image and Video Dataset for Morphological, Histopathological, and Genomic Characterization of Colorectal Polyposis (arxiv.org)

Quantifying Overclaiming Propensity in Frontier LLM Agents (arxiv.org)

RetireOPD: Self-Retiring On-Policy Distillation for Agentic Reinforcement Learning (arxiv.org)

QEncodeBench: Can Large Language Models Encode Classical Problems into Verified Quantum Oracles? (arxiv.org)

Semantic Action Graph: A Shared Representation for Agent Grounding and Human Interpretation of Sports Highlights (arxiv.org)

Large Language Models as Falsifiers for Cyber-Physical Systems (arxiv.org)

HIL-UMI: Bringing Human-in-the-Loop Post-Training of Vision-Language-Action Models to Universal Manipulation Interface (arxiv.org)

Harm Laundering in GPT Models: Evidence That Gender Discrimination Is Transformed Rather Than Reduced Across Safety-Trained Generations (arxiv.org)

GeoAAC: Geometry-Based Adaptive Action Chunking from Denoising Trajectories in VLA Policies (arxiv.org)

TinyCNN: A 193K-Parameter Network for On-Device Plant Disease Detection, with a Cross-Dataset Robustness Diagnosis (arxiv.org)

A Table-Free Index for Tapered Memoization Grids: Compact Out-of-Core Evaluation of Functions of Sorted Arguments (arxiv.org)

Exploring Sparsity and Smoothness of Arbitrary Lp Norms in Adversarial Attacks (arxiv.org)

Chronicle: Cut-Point Replay for Regression Testing of LLM Agents (arxiv.org)

Perturbing the Phase: Analyzing Adversarial Robustness of Complex-Valued Neural Networks (arxiv.org)

A Simulation Platform for AUV Fault Recovery: Exploring LLM-Based Diagnostic Strategies (arxiv.org)

Inference-Engine Fingerprinting Attacks are Practical: Exploring Model-Driven Environmental Discovery, Exploitation, and Escape (arxiv.org)

QUATRO: Query-Adaptive Trust Region Policy Optimization for LLM Fine-tuning (arxiv.org)

Elastic Spectral State Space Models for Train-Once Budgeted Inference (arxiv.org)

Ambient Dataloops: Generative Models for Dataset Refinement (arxiv.org)

Pre-train to Gain: Robust Learning Without Clean Labels (arxiv.org)

Accelerating Visual Policy Learning with Sampling-Based Model Predictive Control (arxiv.org)

Model-Agnostic and Language-Agnostic Voice Pipeline Improvement for the Agriculture Domain (arxiv.org)

Edustories: A Collection of Real-world Case Studies from Classroom Practices (arxiv.org)

greCAPTCHA: Assessing Understanding as Evidence of Research Authorship Under Generative AI (arxiv.org)

A Mathematical Model of Motivated Emotional Mind - Cognitive Embodied System (arxiv.org)

When Do Language-Grounded Explanations Help? A Graph-Bottleneck for Farm Monitoring Interpretable Sheep Facial Pain (arxiv.org)

TouchSight: Bare-Handed Tactile Prediction from Egocentric Video via Generative Visual Augmentation (arxiv.org)

Stress-testing Alignment Midtraining (arxiv.org)

Xeno-Interpretability: Investigating the Alien Minds of LLMs (arxiv.org)

Watermarking Diffusion Language Models (arxiv.org)

A Network Science Approach to Granular Time Series Segmentation (arxiv.org)