AI News

Shared-Roadmap Generation and Evaluator for Multi-Agent Path Planning Using Heterogeneous Graph Neural Network (arxiv.org)

Talking with Language Models (arxiv.org)

Towards AI-Generated Music Plagiarism Detection as a Version Identification Problem (arxiv.org)

U-Space: Uncovering When and Why Uncertainty Arises in Language Models (arxiv.org)

On the Computational Complexity of Hidden Markov Model Identification (arxiv.org)

Same Text, Different Prediction: Serving-Context Nondeterminism in Text Classifiers (arxiv.org)

From Uncertainty to Action: Learning to Steer LLM Agents (arxiv.org)

Workhorse: Learning Robust Whole-Body Humanoid Loco-Manipulation from Human Data (arxiv.org)

CADFather: Autonomous CAD Reconstruction through Coordinated Tool Use (arxiv.org)

Quantize by Drift: Label-Free Mixed-Precision Post-Training Quantization for Text Embedders (arxiv.org)

Sampling SU(N) gauge theory on a 2D lattice from independent plaquettes via holonomies and corner reweighting (arxiv.org)

FreeEvolve: Learning to Evolve Beyond Fixed Loops (arxiv.org)

Few Bits, One Law: Toward W2A4KV2 (arxiv.org)

CM-DPO: Constraint-Margin Direct Preference Optimization for LLM Planning (arxiv.org)

FLoRa: Flight-Assisted Data Collection from Duty Cycling LoRa Nodes under Energy Constraints (arxiv.org)

Correspondences as Decisions: JevNexus for Decision-Centric Schema Matching (arxiv.org)

Q-PACE: Dynamic Precision Allocation for Quantization-Aware Training (arxiv.org)

The Winner's Curse in LLM Self-Improvement Loops: Selection Noise, Lock-in, and Acceptance Rules (arxiv.org)

Beyond Nominal Equilibria: Risk-Averse Multi-Population Mean-Field Games (arxiv.org)

Hardware-aware Calibrated Clustered Attention for Efficient Visual Geometric Transformers (arxiv.org)

SkillRL: Evolving Agents via Recursive Skill-Augmented Reinforcement Learning (arxiv.org)

Node-level Graph Neural Architecture Search Framework (arxiv.org)

ELROND: Exploring and decomposing intrinsic capabilities of diffusion models (arxiv.org)

Twist Flow for Inverse Problems (arxiv.org)

Self-attention summary networks for subsurface velocity-model building from common-image gathers (arxiv.org)

LeCuration: A Tiny World Model as a Data Curation Multi-Tool (arxiv.org)

Lightweight and Versatile Learned Optimization by Recombination of Gradient History (arxiv.org)

Kuration SDK: Addressing the Virtual2Real Gap via Data Curation (arxiv.org)

MovieSTAGE: Scene, Transition, and Global Encoding for Movie-fMRI ADHD Classification (arxiv.org)

Denoising Blocks, Not Tokens: Efficient Compressed Continuous Diffusion with Branching Token Realization (arxiv.org)

Shared Low-rank Basis Factorization for Data-free Mixture-of-Experts Compression (arxiv.org)

Decoupled Optimization for Teacher-Student Semi-Supervised Learning via a Pioneer Student (arxiv.org)

Stability and Diversity of Networked Self-Consuming Generative Ecosystems (arxiv.org)

Multimodal LLMs Can Learn to Read Brain Signals: A Vision--Language Model for Unified Multi-Task EEG Decoding (arxiv.org)

From Retrieval to Customer Context: Evaluating Frontier-Model Systems for Voice-of-Customer Analysis (arxiv.org)

Noise, Denoise, Correct: MCMC Posterior Sampling with Diffusion Priors in Three Steps (arxiv.org)

Collaborative Reasoning Distillation via Cross-Feedback and Coherent Curation (arxiv.org)

LayerRoPE: Dynamic Depth-wise Magnitude & Angular Superposition (arxiv.org)

Shared Geometry As A Rosetta Stone: Cross-Modal Alignment Without Paired Data (arxiv.org)

Self-Consuming Generative Models with Co-Evolving Human Preferences (arxiv.org)

What a Reporting Convention Hides: A Matched-Budget Audit of Quantum Natural Gradient with an Exactly Computed Metric (arxiv.org)

An extended deep energy method for thermo-mechanical crack propagation (arxiv.org)

GeoPrior-Mamba: Structured Process Priors with Mamba for Fine-Resolution XCO2 Reconstruction (arxiv.org)

SoftSEEPS improves ML-based precipitation forecasting (arxiv.org)

When Should an In-Context Learner Expand Its Hypothesis Space? (arxiv.org)

MORA: Modeling Observed Changes for Drift-Robust Time-Series Anomaly Detection (arxiv.org)

CHASE: Channel-Aligned Structure Exploitation for Geometry-Aware Model Engineering (arxiv.org)

Safe on Average, Unsafe in the Tail: When Is the Episodic-Cost Tail Controllable? (arxiv.org)

When Rank Rises as LLMs Degrade (arxiv.org)

Physics-Informed Neural Plasticity: PDE Solvers That Reshape Themselves (arxiv.org)