AI News

A Unified Adaptive Feature Composition Framework for Multi-Task Generalization in Wireless Foundation Models (arxiv.org)

Multi-Level Analyzation of Imbalance to Resolve Non-IID-Ness in Federated Learning (arxiv.org)

When Design Rules Break: Benchmark Composition Determines Whether Label Informativeness Predicts GNN Aggregator Choice (arxiv.org)

DUET -- Dual User Embedding Transformers for Offsite Conversion Prediction (arxiv.org)

Deep Slice Interpolation for Reducing Through-Plane Anisotropy and Noise in Head CT (arxiv.org)

Learning Where to Simulate: Generative Active Sampling for Online PDE Surrogate Training (arxiv.org)

GAGI: A Gini-Adjusted GDP-per-Capita Index for Distribution-Aware Macroeconomic Welfare Monitoring (arxiv.org)

Anomaly Detection and Root Cause Analysis for Microservice Systems (arxiv.org)

Interactions Between Crosscoder Features: A Compact Proofs Perspective (arxiv.org)

Conformal Prediction for Neural Operators: Distribution-Free Uncertainty Quantification in Physics Simulation (arxiv.org)

Co-GLANCE: Uncertainty-Aware Active Perception for Heterogeneous Robot Teaming (arxiv.org)

SPDM: Geometry-Modulated State Space Modeling with Manifold Constraints for Time Series Forecasting (arxiv.org)

IntentKV: Cross-Turn Intent-Aware KV Cache Pruning for Agent Inference (arxiv.org)

RKSC: Reasoning-Aware KV Cache Sharing and Confident Early Exit for Multi-Step LLM Inference (arxiv.org)

The Whale That Outswam Evolution: Swarm Intelligence Maximises Memory in Connectome Reservoirs (arxiv.org)

Less Context, More Accuracy: A Bi-Temporal Memory Engine for LLM Agents Where a Lean Retrieved Context Beats the Full History (arxiv.org)

When Attribution Patching Lies: Diagnosis and a Second-Order Correction (arxiv.org)

HMAF: A Hierarchical Multi-Slot GD-RTB Allocation Framework (arxiv.org)

Tractogram foundation model (arxiv.org)

TENP: Trapezoidal Expert Neuron Pruning For Mixture-of-Experts (arxiv.org)

QSplitFL: Capability Aware Deep Q-Learning for Optimal Split Point Selection in Split Federated Learning (arxiv.org)

SPACE: Source-free Proxy Anchor Concept Erasure for MLLMs (arxiv.org)

Two to Tango: Coupled Task-Reference Selection for Safe LLM Fine-tuning (arxiv.org)

LLM-as-a-Discriminator: When Synthetic Tables Still Look Real (arxiv.org)

Representation Curriculum: Stagewise Training for Robust Ranking and Allocation (arxiv.org)

PreAct-Bench: Benchmarking Predictive Monitoring in LLMs (arxiv.org)

SocraticPO: Policy Optimization via Interactive Guidance (arxiv.org)

SHAPE: Coalition-Aware Expert Pruning for Sparse Mixture-of-Experts LLMs (arxiv.org)

A Controlled Audit of Pretraining Contamination in Public Medical Vision-Language Benchmarks (arxiv.org)

Alignment Collapse Under KV Cache Quantization: Diagnosis and Mitigation (arxiv.org)

CodeAlchemy: Synthetic Code Rewriting at Scale (arxiv.org)

From Confident Closing to Silent Failure: Characterizing False Success in LLM Agents (arxiv.org)

Blurry Window Attention (arxiv.org)

Time Series as Language: A Universal Tokenizer for General-Purpose Time Series Foundation Models (arxiv.org)

Conformal Risk Prediction for Non-Alcoholic Fatty Liver Disease Using Gradient Boosting with Distribution-Free Coverages (arxiv.org)

SoK: Colluding Adversaries in Machine Learning Pipelines (arxiv.org)

Hyperbolic Neural Population Geometry Benefits Computation (arxiv.org)

CANVAS: Captioning Art with Narrative Visual-Audio AI Systems (arxiv.org)

The Interlocutor Effect: Why LLMs Leak More Personal Data to Agents Than Humans (arxiv.org)

An LLM-Native Psychometric Instrument Does Not Predict LLM Behavior: Evidence Across 25 Models (arxiv.org)

Integrated Real-Time Motion Tracking and AI Analysis for Athletic Performance Optimization (arxiv.org)

Failure Modes of Deep Multi-Agent RL in Asynchronous Pricing: Reproducible Triggers, Trace Diagnostics, and a Partial Fix (arxiv.org)

Self-EmoQ: Plutchik-Guided Value-based Planning to Drive Streaming Emotional TTS (arxiv.org)

CollabSkill: Evaluating Human-Agent Collaboration On Real-World Tasks (arxiv.org)

Agentic Social Affordance Framework (ASAF): Agent Identity Design as a Collaboration Interface in Multi-Agent Systems (arxiv.org)

AI-Driven Analytics of Team-Teaching Talk: Acoustic Patterns across Experience, Cohorts and the Learning Design (arxiv.org)

TD-Grokking: Learning from Zero-Reward Problems by Training-Time Decomposition (arxiv.org)

Mitigating Manifold Departure: Uncertainty-Aware Subspace Rectification for Trustworthy MLLM Decoding (arxiv.org)

Uncertainty-aware Multi-fidelity Closure via Conditional Normalizing Flows (arxiv.org)

SynIB: Informational Bottleneck for Maximizing Synergy in Multimodal Learning (arxiv.org)