AI News

Temporal Context Reinstatement Drives Episodic-Like Order Memory in Long-Context Language Models (arxiv.org)

PhononBench-MP40: a spectrum-resolved benchmark dataset for phonon stability (arxiv.org)

Multi-Objective Structured Pruning of LLMs for Latency and Model Size Optimization (arxiv.org)

Kinship Verification through a Forest Neural Network (arxiv.org)

The Scaffold Effect in Coding Agents: Harness Choice as a Hidden Variable in Coding-Agent Evaluation (arxiv.org)

TriSP: Tri-Signal Structured Pruning for Large Language Models (arxiv.org)

SimBEV2X: A Large-Scale Dataset and Data Generation Tool for Multi-Task Vehicle-to-Everything Cooperative Perception (arxiv.org)

Long-Tailed Medical Image Classification (arxiv.org)

BERT-based Models vs. Large Language Models for Low-Resource Named Entity Recognition: A Comparative Study on Marathi (arxiv.org)

A Frozen 12B Beats Frontier Models on Verified Work: 100% Accuracy, 0 Tokens, Bit-Exact, Forever (arxiv.org)

Attribution and Uncertainty Behavior of Learned Residual Gyro Correction for Gyro-Stellar Estimation (arxiv.org)

DP-IVON-Gradsq: Differentially Private Squared-Gradient Improved Variational Online Newton (arxiv.org)

Agent-UCT: Upper Confidence Bounds Applied to Trees for Agentic Workflow Optimization with Cost-Awareness (arxiv.org)

TRE: Training-Free Hallucination Detection for Diffusion Language Models (arxiv.org)

MIITA: Memory-Induced Inference-Time Adaptation for Continual Learning with Small Language Models (arxiv.org)

DeepLens Diagnosis Agent: Agentic Workflow Design Lets a Small Reasoning Model Compete with Frontier LLMs (arxiv.org)

Causal-TS: A Python Library for Causal Discovery in High-Dimensional and Nonstationary Time Series (arxiv.org)

Loss-Aware Feature-Map Pruning in Convolutional Neural Networks Using Multi-Armed Bandits (arxiv.org)

CodexGraph: Bridging Large Language Models and Code Repositories via Code Graph Databases (arxiv.org)

MolCryst-MLIPs: A Machine-Learned Interatomic Potentials Database for Molecular Crystals (arxiv.org)

AI-Assisted Causal Inference and Mediation Analyses of Environmental and Psychosocial Determinants of Subjective Cognitive Difficulties in the All of Us Research Program (arxiv.org)

CALMRec: Causally Aligned Language Memory for Long-Horizon Recommendation (arxiv.org)

An Unofficial FastLAS Tutorial: A Programmer's Guide (arxiv.org)

TRACE-CTI: Auditable Post-Extraction Governance of TTP Claims with Knowledge Graphs (arxiv.org)

Generative Artificial Intelligence (GenAI) to convert images of queuing networks into verifiable simulation models: an open-weight LLM workflow approach (arxiv.org)

Epistemic Norms for AI Safety and Alignment Research (arxiv.org)

To Erase, or Not to Erase: Robust Training-Free Concept Erasure with Preservation aware Adaptive Ranked Subspace Expansion (arxiv.org)

Action from Adjacent Set in Physical Space Outperforms the Best Prediction in World Models (arxiv.org)

Plans Work in Mysterious Ways: Evaluating a Plan Mode for Spreadsheet Agents (arxiv.org)

Failures Reveal What Metrics Miss: An Evidence-Driven Agent for Recursive Refinement of ECG Classifiers (arxiv.org)

ATLAS: Automated Approximation of Transformers for Efficient Homomorphic Inference in One Hour (arxiv.org)

Variational-Ising-Attention (VIA):TailoredAttentionMattersfor Science (arxiv.org)

Stacking the Deck: Tunable Trainability in Stacked LCUs (arxiv.org)

Simulating Tenant Responses to Energy Policy Interventions with Transaction-Cost-Aware LLM Age (arxiv.org)

Approximate reservoir computing with a semiconductor laser for reducing energy consumption (arxiv.org)

LLM-SoccerArena: Benchmarking LLMs on Real-World Predictions in Sports (arxiv.org)

MIME: Multimodal Interactive Motion Encoder (arxiv.org)

Training Language Models to Cooperate with Inference-Time Controllers (arxiv.org)

DSCH-Loss: A Dynamic Semantic Channel Objective for Deep Semantic Hashing (arxiv.org)

Visual Token Compression Enhances Robustness of MLLMs (arxiv.org)

A New Kind of Adversarial Example: Measuring the Human-Model Gap, and Its Relationship to OOD Detection (arxiv.org)

Structure over Depth: A Single-Block Spatio-Temporal Transformer for Multi-Entity Reasoning (arxiv.org)

Obliviate: Efficient Unlearning in Recommender Systems (arxiv.org)

Frustratingly Simple Black-Box Adaptation of Language Models via Logit Bias (arxiv.org)

AIR-BENCH Live: An Evolving Safety Benchmark for Foundation Models (arxiv.org)

MarineEVT: Advancing Event-Centric Marine Video Understanding via Visual Tool Reasoning (arxiv.org)

CAPT: A Multi-task Continuous Autoregressive Transformer enabling Cross-dataset and Cross-species Transfer for Calcium Population Dynamics (arxiv.org)

MS-GPT: Rethinking MS/MS De Novo Structure Elucidation as Spectrum-Induced Posterior Querying of a Molecule-Language Model (arxiv.org)

Extreme Volatility Warning under Label Scarcity via Multi-Source Anomaly Fusion (arxiv.org)

Reason Popper-ly: Patching In-Context Reasoning with Inductive Logic Programming (arxiv.org)