AI News

LILA: Calibration-Free Structured Pruning of Large Language Models via Latent Spectral Geometry (arxiv.org)

Valerant: An Automatic Navigable Game Map Generator via Action-Conditioned World Model Exploration (arxiv.org)

Thompson Sampling for Non-Monotone Convex Ridge Bandits: Monotonicity Is Not Needed for Polynomial Regret (arxiv.org)

Beyond Solver Verdicts: Generative Reward Models for Autoformalization (arxiv.org)

3D Point Splatting for mmWave Radar Novel View Synthesis (arxiv.org)

Teacher Geometry Shapes Learnability in Teacher-Student Networks (arxiv.org)

Decision-Focused Active Learning for Scale-Aware Critical-Materials Recovery (arxiv.org)

Hologram Representation via Quadratic Phase Gaussian Splatting (arxiv.org)

HiRAD: A Flexible Large-Scale AGV Routing System (arxiv.org)

EGGROLL, Unrolled: Understanding and Improving Low-Rank Evolution Strategies at Scale (arxiv.org)

Explainability Assistant: A Conversational XAI Interface for Interpreting Energy Consumption Models (arxiv.org)

Enabling Knowledge Graph Understanding at Scale with the EXplore Your Graphs ENgine (EXYGEN) (arxiv.org)

Procedural Memory Under Change: Reuse and Interference in Controlled Web Tasks (arxiv.org)

Domain-Specific Hallucination Detection in Large Language Models (arxiv.org)

Building py-kvcache: A Performance Characterization of External KV Caching for vLLM with NVMe SSDs (arxiv.org)

New Evidence, Same Choice: Testing Physical Experiment Selection in Vision Language Models (arxiv.org)

Data Scarcity and Model Sparsity: Mixtures-of-Experts Overfit More to Repeated Data (arxiv.org)

Active noise cancellation on open-ear smart glasses (arxiv.org)

OntologyAligner: Ontology-Aligned Retrieval and Hierarchy-Guided Large Language Model Reranking for Biomedical Ontology Normalization (arxiv.org)

RiLM: Parameter-Efficient Language Modeling via Geodesic Decoding (arxiv.org)

Toward a First-Principles Update Geometry for the Language-Model Head (arxiv.org)

The information geometry of large language models is shared, learned, and controllable (arxiv.org)

Understanding LoRA Rank Trade-offs in Diffusion Model Fine-Tuning (arxiv.org)

SG-Blend: Learning an Interpolation Between Improved Swish and GELU for Robust Neural Representations (arxiv.org)

Time-Frequency Geometric Cross-Attention for Chunked Vision-Language-Action Models (arxiv.org)

Structural Process Supervision for Latent Chain-of-Thought Reasoning (arxiv.org)

Belief-State Engine: Augmenting LLMs for Principled Planning Under Partial Observability (arxiv.org)

Learning Intrinsic Water-Quality Dynamics with Rainfall for Data-Driven Forecasting (arxiv.org)

What Should an Agent Forget? Separating What Is Stored from What Is Used (arxiv.org)

An Autonomous GeoAI Agent for Arctic Eco-Navigation (arxiv.org)

Decision Transformer for UAV-Mounted RIS-Assisted Dynamic D2D Communications (arxiv.org)

Efficient Leakage-Free Neural Architecture Search under Leave-One-Subject-Out Evaluation (arxiv.org)

Incentives to Offer Algorithmic Recourse (arxiv.org)

ObstaDiff: Generalizable Diffusion Policy Learning via Obstacle-aware Representations (arxiv.org)

BiHDTrans: binary hyperdimensional transformer for efficient multivariate time series classification (arxiv.org)

CertDW: Towards Certified Dataset Ownership Verification via Conformal Calibration (arxiv.org)

Proof-Carrying Cognition: Closing the Verification Gap with Reality-Settled Reward (arxiv.org)

Hyperbolic Geometry for Open-World Object Detection in Remote Sensing Imagery (arxiv.org)

Structural priors for data-efficient language learning (arxiv.org)

CARTS: Contextual Autoregressive Rank Transcoding Steganography for Full-Capacity Keyed Text Encoding (arxiv.org)

Generalization Analysis of Distributed Kernel-based Robust Gradient Descent Algorithms (arxiv.org)

RAP: Research Attention Prediction Reveals Target-Conditioned Evidence Acquisition Biases (arxiv.org)

How Fragile Is Safety Alignment at Frontier Scale? A Single-Direction Attack on a 320B MoE (arxiv.org)

CS-Guard: Benchmarking LLM Guardrails for Code Generation Security (arxiv.org)

Grounded Evaluation and Repair for NL-to-PDDL Problem Generation (arxiv.org)

The Vibe Shift in Software Engineering: Evaluating AI-Led Conversational Programming for Performance, Cognition, and Responsible Adoption (arxiv.org)

MUC-FL: Block-Wise Marginal Utility Contribution for Communication-Efficient Federated Learning (arxiv.org)

Rethinking Radiomap Blind Prediction with Limited Environment and Configuration Representations (arxiv.org)

TailProp: content-adaptive light- and heavy-tailed propagation for vision (arxiv.org)

PEARL: A Task-Aware Framework for Evaluating Differentially Private Synthetic Educational Data (arxiv.org)