AI News

SimpleDesign: A Joint Model for Protein Sequence and Structure Codesign (arxiv.org)

Time Without Timesteps: Simulating Coupled Dynamical Systems via Self-Consistency (arxiv.org)

Gradients Know What Outcomes Don't: Unlocking Reinforcement Learning for LLM Reasoning with Gradient-Aligned Rewards (arxiv.org)

A Prompt-Engineering Approach to Develop Scalable, Flexible, and Real-Time Hybrid Micro-Level Personalization in a General Purpose AI Teaching Assistant (arxiv.org)

Fresh Memory, Stale Plans: Dependency-Scoped Validation for Distributed LLM-Agent Memory (arxiv.org)

Speculative Macro Commit for Faster Tool-Using Agents (arxiv.org)

Counterfactual Routing Using Integer Programming with Constraint Generation (arxiv.org)

Dude: A Dual-Detection Multi-Agent System for Paper-Code Discrepancy Detection (arxiv.org)

A Large Open Multi-Energy Corpus of Soil Compaction Tests, with Machine-Learning Baselines (arxiv.org)

DE-Venus: A Data-Efficient RLVR Framework for Large Language Models (arxiv.org)

Risk and Anomaly Identification for Distribution Network Optimal Operation Based on Reinforcement Learning and Uncertainty Quantification (arxiv.org)

Geometry-Aware Graph Construction via Adaptive Spectral Bandwidth Control (arxiv.org)

Latent Energy Action Planning with World Models (arxiv.org)

Improving precipitation forecasts in an AI weather model using observational data (arxiv.org)

Selective Hypergraph Refinement for Frozen Graph Clustering (arxiv.org)

FlowBalance: Verifier-Grounded Self-Improvement from On-Policy Reasoning Experience (arxiv.org)

B2B Customer Conversion Prediction: A Document Representation, Graph Theory, and CatBoost Driven Methodology (arxiv.org)

The 2026 PNPL Competition: Word Classification and Efficient Cross-Subject Generalisation in LibriBrain100 (arxiv.org)

Language-encoded network topology enables large language models to reason about complex networks (arxiv.org)

Portable Causal Fairness Across Synthetic Data Generator Families (arxiv.org)

Frontier LLMs are effective batch optimizers: Assessing reasoning models in continuous and discrete settings (arxiv.org)

Routing Is Not Enough: Diagnosing Intra-Adapter Subspace Contention in MoE+LoRA Fine-Tuning (arxiv.org)

Kernel Reboot: Breaking the Boundaries of Neural Tangent Kernels for Neural Fields (arxiv.org)

MasterControl Seventeen Every Time (arxiv.org)

Scaling Laws, Tabular Data and Actuarial Ratemaking Models (arxiv.org)

Distilling deep optical flow stereo methods to retrieve dense three-dimensional wind fields (arxiv.org)

The Gradient Does Not See Rank: Rank-Indifference in Matrix-CODI on ProsQA (arxiv.org)

LeanStream: A Speculate-and-Refine Streaming Framework for Efficient on-Device LLM Inference (arxiv.org)

ObserverBench: Testing Mechanistic Estimates for Intervention and Control (arxiv.org)

Verify Before You Distill: Prompt-Level Teacher Gating for On-Policy Distillation (arxiv.org)

Mesh-Native Physics-Informed Graph Surrogates for TCAD-in-the-Loop Design Space Exploration (arxiv.org)

Modern Transformers Are Implicit Hybrids: From Functional Differentiation to Principled Hybrid Architecture Design (arxiv.org)

From Euclidean to Graph-Structured Data: A Survey of Collaborative Learning (arxiv.org)

Learnable composition for neural operators (arxiv.org)

Equation Recast for Canonical Operator Learning Across Parametric PDEs (arxiv.org)

ALRA: Adaptive Local Relational Alignment for Logit-Based Pre-training Distillation of Autoregressive Language Models (arxiv.org)

Entropy-Generated Attention Beyond Softmax and Entmax: Kaniadakis and Reciprocal-Symmetric Abe Operators (arxiv.org)

Distributional Treatment Effect Transportability across Heterogeneous Sites (arxiv.org)

Online Learning of Functional Principal Component Analysis for Multidimensional Functional Data (arxiv.org)

Active learning for data-driven reduced models of parametric differential systems with Bayesian operator inference (arxiv.org)

Reliable Selection of Heterogeneous Treatment Effect Estimators (arxiv.org)

Cooperative Multi-Task Semantic Communication for Joint Classification and Regression Tasks (arxiv.org)

High-Dimensional Learning Dynamics of Attention-Indexed Models (arxiv.org)

Q-Edge: Symmetry-Reduced Quantum Simulation of Structured Extreme Dependence (arxiv.org)

Correlated initialization of deep residual networks (arxiv.org)

Semiparametric Inference for Counterfactual Regression under Intervention-Driven Shift (arxiv.org)

Spectral characteristics of autoencoder parameters as a vector representation of data (arxiv.org)

How Perturbations Propagate: A Multi-Level Analysis of Robustness in Large Language Models (arxiv.org)

Discrete Gromov-Wasserstein Duality: Algorithms and Isomorphism Testing (arxiv.org)

Unifying Conformal Language Tasks with In-Context Ensembles (arxiv.org)