AI News

CAPruner: Conceptual-Adjacent Scene Graph Pruner for Enhancing 3D Spatial Reasoning of Large Language Models (arxiv.org)

Community-Specific Slang and Entity Detection via Semantic Shift in Fine-Tuned Language Models (arxiv.org)

TinyJudge: Unverifiable Constraint Alignment via Lightweight Specialist Ensembles (arxiv.org)

VisualLeakBench: Reproducible Action-Boundary Propagation Failures in Vision-Language Agents (arxiv.org)

A Mechanistic Analysis of Adversarial Fine-tuning of Vision Transformers (arxiv.org)

SlideCheck: Guiding Self-Supervised Pretraining of Pathology Foundation Models via Dataset Distributions (arxiv.org)

Multimodal Group Emotion Recognition In-the-Wild Towards a Privacy-Safe Non-Individual Approach (arxiv.org)

An Agency-Transferring Model-Free Policy Enhancement Technique (arxiv.org)

Rethinking the Divergence Regularization in LLM RL (arxiv.org)

Topological Neural Operators (arxiv.org)

Zero Touch Predictive Orchestration: Automating Time-Series Models for the Cloud-Edge Continuum (arxiv.org)

iOSWorld: A Benchmark for Personally Intelligent Phone Agents (arxiv.org)

Preserving Plasticity in Continual Learning via Dynamical Isometry (arxiv.org)

Disentanglement with Holographic Reduced Representations (arxiv.org)

A Systematic Study of Behavioral Cloning for Scientific Data Annotation (arxiv.org)

SurfDesign: Effective Protein Design on Molecular Surfaces (arxiv.org)

Phantom transitions in language model fine-tuning (arxiv.org)

Page image classifier fine-tuned on century-spanning archives of scanned documents for further content-specific processing (arxiv.org)

Selecting New Measurement Locations to Diversify Traffic-Pattern Coverage: A Real-World Evaluation for Total Traffic Volume Estimation (arxiv.org)

Symbolic Reasoning Frameworks Modulate LLM Risk Aversion in Multi-Agent Strategic Settings (arxiv.org)

Considerations for an Integrated Detector Design at FCC-ee: A Human-AI Exploration (arxiv.org)

The Montparnasse Algorithm for RNA Design (arxiv.org)

Liberating LLM Capabilities in Full-Duplex Speech Models (arxiv.org)

Beyond Item IDs: Scaling Short-Form-Video Recommendation via Semantic-Native Long Sequence Modeling (arxiv.org)

MetaEvo: A Meta-Optimization Framework for Experience-Driven Agent Evolution (arxiv.org)

Sample-Efficient Post-Training for LEGO Spatial-Physics Reasoning (arxiv.org)

LFNO: Bridging Laplace and Fourier via Transient-Steady Decomposition (arxiv.org)

Reachability and asymptotics of Gaussian Transformer dynamics (arxiv.org)

SRT: Super-Resolution for Time Series via Disentangled Rectified Flow (arxiv.org)

DiffoR: A Unified Continuous Generative Framework for Universal Ordinal Regression (arxiv.org)

A Topological Characterization of Graph Neural Networks via Stochastic Block Model Embeddings on the n-Sphere (arxiv.org)

Repetition Mismatch: Why Data Mixture Experiments Don't Scale and How to Fix Them (arxiv.org)

Shortcuts in the Tail: Debiasing via Post-Hoc Spectral Compression of Fine-Tuning Updates (arxiv.org)

UNIQ: Conformal Calibration for Adaptive Conservatism in Offline Reinforcement Learning (arxiv.org)

Optimality of Sequential Filtering Under Independent Cost and Selectivity Models (arxiv.org)

The Routing Plateau: Understanding and Breaking the Accuracy Limits of LLM Routers (arxiv.org)

From Human Guidance to Autonomy: Agent Skill System for End-to-End LLM Deployment on Spatial NPUs (arxiv.org)

ResearchClawBench: A Benchmark for End-to-End Autonomous Scientific Research (arxiv.org)

Outage Detection in Self-Healing Smart Grids Using Reinforcement Learning with Spectral Graph Neural Networks (arxiv.org)

Eyes All Around: Design and Analysis of 360-Degree LiDAR Perception Using Equivariant Feature Learning in Unstructured Traffic (arxiv.org)

Structured Neuron Pruning in Deep Neural Networks Using Multi-Armed Bandits (arxiv.org)

Can You Trust What You See? Human and AI Detection of Synthetic Legal Evidence (arxiv.org)

BEACON: Behavioral Entropy Aggregation for Cross-Model Hallucination Detection in Large Language Models (arxiv.org)

Pointwise Complexity for Gaussian Fields: Upper Envelopes, Algorithmic Lower Bounds, and Separation (arxiv.org)

GraphLoRA: Structure-Aware Low-Rank Adaptation for Large Language Model Recommendation (arxiv.org)

Implicit Causal Graph Construction in Text via Chain Discovery (arxiv.org)

ABLE: Representing and Mapping LLMs via Attribution-Based Large-model Embedding (arxiv.org)

Retrieval Augmented Generation Framework for the Nepali Legal Domain Question Answering (arxiv.org)

Evaluating Hallucinations in Domain-Adapted Large Language Models (arxiv.org)

A retrieval conditioned rebinding circuit for dynamic entity tracking in large language models (arxiv.org)