AI News

Mesh-Native Tokenisation: Retrofitting Foundation Models to Unstructured Data (arxiv.org)

Self-Preservation in Language Models (arxiv.org)

WoundAmbit: Bridging State-of-the-Art Semantic Segmentation and Real-World Wound Care (arxiv.org)

DuoRate: Two-Rate Per-Expert Precision Allocation for Mixture-of-Experts Serving (arxiv.org)

SBNO : Schr\"odinger Bridge Neural Operator for the Forward and Inverse Problems under Missing Information (arxiv.org)

Sustainable Cloud-Fog-Edge Systems for Smart Buildings: Algorithmic Adaptation under Memory Constraints and Dynamic Orchestration Strategies (arxiv.org)

Structure-Aware Numerical Stabilization for Half-Precision Proximal Policy Optimization (arxiv.org)

MORAD: Reinforcement Learning with Multi-Objective Rewards for RNA Inverse Design (arxiv.org)

PDE-PFN: Prior-Data Fitted Neural PDE Solver (arxiv.org)

CopRoute: Dependence-Aware Conformal Routing for Language Models (arxiv.org)

Data-Adaptive Mahalanobis Metric Learning for Cross-Head Attention in Transformers (arxiv.org)

Neural Surrogate Modeling of Ultrasonic Wave Propagation in Heterogeneous Media (arxiv.org)

A Multi-Agent LLM System for Automated Tuning of Specialized Neural Network Models (arxiv.org)

Recovering Irreducible Three-Body Interactions with Lagrangian Neural Networks (arxiv.org)

Learning Size Independent Constraint Models from Input Data and Positive and Negative Example Solutions (arxiv.org)

Humanize Olympiad Agents: A Multi-Session Workflow for Olympiad Problem Solving with Solve, Review, Wiki, Verify (arxiv.org)

MENTOR: Learning from Failure with On-Policy Pedagogical Intervention (arxiv.org)

Constructing Pareto Frontiers of Optimal Reasoning Termination Methods for Overthinking in Reasoning Models (arxiv.org)

CRM: Retrieval Model with Controllable Condition (arxiv.org)

MAGMA-Bench: Evaluating Long-Horizon Embodied Agents under Dynamic Tasks and Execution Failures (arxiv.org)

Beyond Domain-Routed Supervision: Cross-Teacher Guarding for Multi-Teacher On-Policy Distillation (arxiv.org)

Causal Effect Identifiability in Causal Abstractions: A Strict Hierarchy of Criteria (arxiv.org)

Retrieval-Augmented Multimodal Language Model with Heterogeneous Evidence Integration for Clinical Prognosis (arxiv.org)

Optimal Algorithms for Nonsmooth Nonconvex Stochastic Optimization with Constraints (arxiv.org)

Tabular Foundation Models Are Effectively Shallow (arxiv.org)

State-OPD: State-Grounded On-Policy Distillation for Multi-Turn Agents (arxiv.org)

Learning Schwarzschild Geodesic Motion with Lagrangian Neural Networks (arxiv.org)

BLEND: Balancing Personalization vs. Generalization in Federated Vision-Language Models (arxiv.org)

Humanize MLE Agents: Multi-Agent Session Alternation Beyond Single-Agent Limits (arxiv.org)

Branching On-Policy Distillation (arxiv.org)

A Local Agentic AI Framework to Generate Edge Case Scenarios for Safety Assurance (arxiv.org)

SqueezeBench: Saturation-Resilient Evaluation of Mathematical Capability (arxiv.org)

How Much Computation Does Every Forecast Need? Structured Block Skipping in Two Time-Series Foundation Models (arxiv.org)

ReSFlow: Rectifying Categorical Flows on Statistical Manifolds for One-Step Generation (arxiv.org)

Tensor-based Fusion of Multi-modal Data for Predicting Cancer Risk Levels (arxiv.org)

Monarch-MoE: Structured Experts for Multimodal Continual Instruction Tuning (arxiv.org)

Agentic Large Language Models (arxiv.org)

Rollout-Time Intervention: Unlocking Agentic Reinforcement Learning Beyond Policy Limits (arxiv.org)

AI Models Can Provably Hide Arbitrary Capabilities (arxiv.org)

Rethinking Relative Policy Differences Transfer of On-Policy Distillation (arxiv.org)

S$^3$R: Fast and Faithful RL Rollouts via Self-Sparse Speculative Decoding (arxiv.org)

Does Learning Recurrent Networks Matter in Partially Observable RL? (arxiv.org)

WaSA: Wavelet Scattering Autoencoders for Unsupervised Offline Detection of Malicious Clients in Federated Learning (arxiv.org)

Inference-Time Steering Across Scaffolds and Epitopes for Sample-Efficient De Novo Nanobody Design (arxiv.org)

Local Diagnostics of Learned Event Representations for Anomaly Detection at the HL-LHC (arxiv.org)

Web$hell: A Unified Interface for Agentic Web Access (arxiv.org)

RegulaTutor: A Theory-Informed Framework for Long-Term Personalized Learning (arxiv.org)

SWE-Critix: Post-Training Generalist Verifiers for Assessing Coding-Agent Trajectories (arxiv.org)

Continually Post-training Small Models via Synthetic Self-distillation (arxiv.org)

Scaling RNA Foundation Models through Iterative Refinement (arxiv.org)