AI News

Adaptation Interfaces for In-Context Tabular Foundation Models in Time-to-Event Prediction (arxiv.org)

From Deep to Shallow: Unconstrained and Efficient Layer Merging Strategy (arxiv.org)

When Genomic Masking Priors Fail to Transfer: Strong Variant Prediction, Weak Functional Generation (arxiv.org)

KVMem: Virtualizing Million-Token Agent Workspaces on a Consumer GPU (arxiv.org)

Uncensored Open-weight Models: Redistribution as the Persistence Layer (arxiv.org)

PACE: Propagation-Aware Collaborative Correction for One-Shot Personalized Federated Graph Learning (arxiv.org)

Communication-Efficient Personalized Federated Learning via Layer-Wise Multi-Threshold Random Sketching (arxiv.org)

Federated Attack Campaign Detection via Contrastive Encoding of Threat Indicators in Gradient Updates (arxiv.org)

BeaconKV: Key-Value Cache Compression Guided by Beacon Queries for Efficient Large Reasoning Model Inference (arxiv.org)

How Faithful Is Attribution for Sales Forecasting? A Counterfactual Study (arxiv.org)

Learning-Augmented Algorithms: Guarantees, Construction Mechanisms, and System-Level Implications (arxiv.org)

A Robust Watermark-based Fingerprint Framework for GNNs Ownership Verification (arxiv.org)

Resilience Beyond Stationary Client Unavailability: Unlocking Efficient and Unbiased Federated Learning (arxiv.org)

A Fairness Audit of the Duckworth-Lewis-Stern Method: Format-Specific and Gender-Differential Bias, with an Interpretable Calibration Layer for Cricket Target Revision (arxiv.org)

Spectral-Target Physical Latent Structuring for JEPA-Style World Models (arxiv.org)

Training Large Language Models for Small-Molecule Design with Synthetic Task Scaling (arxiv.org)

Locating and Steering Refusal Beyond Attention (arxiv.org)

WEECFP-SuRGE: Wide Embedded Extended Connectivity Fingerprint with Substructure Rotary Graph-distance Encoding (arxiv.org)

SMILE: Bridging Continuous Optimization and Discrete Symbolic Recovery (arxiv.org)

Too Rare to Learn: Prescribed Cyclone Tracks Degrade a Bay of Bengal Ocean Emulator (arxiv.org)

GNN-Guided Graph Coarsening and Adaptive QUBO Penalties for the Capacitated Vehicle Routing Problem with Time Windows on a Quantum Annealer (arxiv.org)

Representation Redundancy and Structural Complexity in Finite-Field Inversion (arxiv.org)

Optimizer Memory Schedules for Outscaling the Overtraining Axis (arxiv.org)

Fast Surrogate Modeling of Excitable and Oscillatory FitzHugh-Nagumo Dynamics with Parametric Neural Operators (arxiv.org)

Mitra-v2 Technical Report (arxiv.org)

Distilled Continuous Diffusion Language Models Can Write Code in Few Steps---or One (arxiv.org)

An Energy-Based Conservative-Dissipative Latent Neural Evolution Operator for Magnetization Dynamics (arxiv.org)

Nested Inductive Bias Framework for SPD Manifold Learning (arxiv.org)

A Repeated-Measurement Study for Cultural Analytics of English Song Lyrics Using Five Large Language Models (arxiv.org)

Embedded Graph Flows for Categorical Graph Generation (arxiv.org)

Beyond a Universal Forecasting Selector: Demand-Conditioned Model Selection across Demand Patterns and Horizons (arxiv.org)

Disentangling Attention in Deep Operator Learning: A Controlled Study of Data-Driven and Physics-Informed Architectures (arxiv.org)

On the Abundance of Critical Points of the t-SNE Energy (arxiv.org)

Conformity Breaks Conformal Prediction (arxiv.org)

A Quantum Variational Approach to Prototypical Recurrent Unit (arxiv.org)

SharedSAE: One Feature Dictionary Across Language Models (arxiv.org)

Modular Deep Recurrent Neural Network: Application to Quadrotors (arxiv.org)

Data-Driven Learning of Unknown Nonlinear Differential Equations Using Functional Analysis (arxiv.org)

BER-PEF: Unified Human Mobility Predictability Evaluation via Bayes Error Rate Estimation (arxiv.org)

Quantum-Assisted Memory-Efficient Training for Parameter-Intensive Wi-Fi-Based Human Activity Recognition (arxiv.org)

A Data Fusion Framework for Grounding Aerospace Surrogate Model via Experimental Wind-Tunnel Observations (arxiv.org)

ProToMEx: Rapid, Interpretable Explanations via Structured Representations (arxiv.org)

Can Activation Steering Capture Multidimensional Authorship Style? (arxiv.org)

A Removal Based Approach to Improve LLM Faithfulness at Test-Time (arxiv.org)

Iris: Climbing to the Search Frontier (arxiv.org)

Data-Optimized Contingency Screening: A Machine Learning Approach to Power System Security (arxiv.org)

Harbor Adapters and Harbor-Index: Infrastructure and a Curated Meta-Dataset for Large-Scale Agentic Evaluation (arxiv.org)

From Matching Models to Recruiting Agents: A Systematized Narrative Review of AI Recruitment Systems, Evaluation, and Governance (arxiv.org)

EXAONE Forecast for Finance (arxiv.org)

Reinforcement Learning for improving Large Language Models' Catalan text simplification capabilities (arxiv.org)