AI News

Efficient Reasoning via Constrained Optimization in Latent Space (arxiv.org)

FlowState: Execution State as Memory for Long-Horizon LLM Agents (arxiv.org)

Visible Reasoning Is Not a Universal Optimizer: Persona- and Thinking-Dependent Effects in Analytics Code Generation (arxiv.org)

Coverage-Aware Reasoning with Medical Tokens for Diagnosis Prediction (arxiv.org)

Leakage-Controlled Multimodal Learning for Diagnosis and Progression Prediction in Alzheimer's Disease Research (arxiv.org)

Beyond the Ergodic Wall: A Discrete Geometric Physics Sandbox for Analysing AI Scaling Limits and Complexity Collapse (arxiv.org)

PXtal: Learning to Align Powder X-Ray Diffraction and Crystal Structures under Information Asymmetry across Modalities (arxiv.org)

Beyond the Remembered World: Predictive 4D Belief for Persistent Navigation in Evolving Worlds (arxiv.org)

Growing an Agent/Prover Interface: Evolutionary Tool Design for Cost-Efficient Theorem Proving in Rocq and Lean (arxiv.org)

Empty Commitments: When Agents Promise What They Cannot Deliver (arxiv.org)

RealCompanion: Benchmarking Human Understanding from Reasoning over Longitudinal Real-World Conversations (arxiv.org)

Beyond Owls: Subliminal Learning Can Transfer Learned Capabilities and Backdoors (arxiv.org)

Learning infinite context windows in recurrent architectures via spatial neural computing (arxiv.org)

Conversational Task Disambiguation over Tabular Data: Leakage-Aware Formulation, Benchmark Suite, and Training (arxiv.org)

NEMORA: Neural Equivariant Multipole Operators for Long-Range Atomistic Learning (arxiv.org)

MemoWM: How World Models Change What Agents Need to Remember (arxiv.org)

How Many Repeated Pairwise Comparisons Are Needed for Ranking under Heterogeneity? (arxiv.org)

Symbolic Density Estimators for Unnormalized Distributions (arxiv.org)

Controlled Acquisition and Abstention in Three-Channel Score Conflicts (arxiv.org)

Evaluating Rubric Generation with Interventional Transfer (arxiv.org)

MemTrace: State-Consistent Memory for Long-Horizon Coding Agents (arxiv.org)

How Should a Prompt Optimizer Spend a Tight Budget? BudgetAPO with Noise-Adaptive Evaluation (arxiv.org)

The Review Lottery: Benchmarking an Observational Estimator of Peer-Review Noise (ICLR 2017-2025) (arxiv.org)

Humanize: Judgement Engineering for Agentic Coding (arxiv.org)

Humanity's Sixth Sense: Benchmarking Intuitive Visual Reasoning in Multimodal Models (arxiv.org)

MIMESIS: Learning User Simulators as Training Environments for Interactive Agents (arxiv.org)

AgentTime: Can Agents Estimate and Control Their Own Runtime? (arxiv.org)

Policy Learning with a Language Bottleneck (arxiv.org)

Multi-Bandwidth Distribution Matching Distillation: On the Equivalence of Distribution Matching Distillation and Drifting Models (arxiv.org)

MotherTree: Meta-learning on synthetic data improves decision tree training (arxiv.org)

Amortized Off-Policy Evaluation for LLMs (arxiv.org)

Similar Predictive Fit but Different Latent Dynamics: Characterizing Learned Dynamical Structure in Personalized Models of Brain Disorders (arxiv.org)

Policy Alignment: New Signals for Membership Auditing in On-Policy Distillation (arxiv.org)

MotiveMob: Motivation as Semantic Action for Closed-Loop Human Mobility Generation (arxiv.org)

Zatom-2: Multitask Pretraining on Atomistic Data for Generative Modeling across Domains (arxiv.org)

Generative Adversarial Loops (arxiv.org)

Compactness and Consistency: A Conjoint Framework for Deep Graph Clustering (arxiv.org)

When to Intervene? State-Aware Sparse Manipulation in Federated Reinforcement Learning (arxiv.org)

Conditional Transfer from Controlled Pretraining Mixtures to Code (arxiv.org)

Best of Both Worlds in Federated LSA: Speedup When Possible, Personalization Always (arxiv.org)

Sera: Semantic Representation Aggregation for Reliable and Interpretable Battery Health Forecasting (arxiv.org)

Uncertainty-Aware Optimization for Physics-Aware Highway Trajectory Prediction (arxiv.org)

BEAT: Balanced Frequency Adaptive Tuning for Long-Term Time-Series Forecasting (arxiv.org)

Heterogeneous-Modal Unsupervised Domain Adaptation via Latent Space Bridging (arxiv.org)

CAF\'E: Causal Black-Box Testing of Machine Unlearning (arxiv.org)

AyurParam: A State-of-the-Art Bilingual Language Model for Ayurveda (arxiv.org)

Attention when you need (arxiv.org)

Constructing Structured Decision Sources for Consensus-Based Pseudo-Label Learning (arxiv.org)

Beyond Action Entropy: Quotient-Space Exploration for Genome-Scale Metabolic Model Repair (arxiv.org)

Evi-VN: Hard Region Guided Virtual Node Evidence Injection for GNN-Based Fraud Detection (arxiv.org)