AI News

Political Narratives in the 2024 Moldovan Presidential Election: Stance, Amplification, and Influence (arxiv.org)

DriveWorld3D Evidence Constrained 3D World Modeling and Generative Simulation for Real-World Autonomous Driving (arxiv.org)

Recoverability-Controlled Safety Miscalibration in Large Language Models (arxiv.org)

Measuring Mechanistic Uncertainty in Large Language Models via Causal Robustness (arxiv.org)

Really Need Recursion? Revisiting One-pass Registration with Multi-level Multi-space Contextual Guidance (arxiv.org)

From Individual Songs to Music Corpora: Evaluating Explainable Harmonic Decision-Making on HookTheory (arxiv.org)

Towards Adaptive and Real-Survey-Aligned Simulated Urban Mobility Behaviour of Generative Agents (arxiv.org)

Eliciting Specifications: When Does Interactive Clarification Help? (arxiv.org)

AI-Driven Reproducibility and Analytical Robustness in Scientific Discovery (arxiv.org)

Self-Building Labs (arxiv.org)

Beyond SAEs: Consistent Concept Discovery with Sparse Coding (arxiv.org)

FuTransformer: FP8 prefill and INT4 decode for Llama-style inference on Hopper (arxiv.org)

DrugHarness: An Evolvable Agent Harness for Computational Drug Discovery (arxiv.org)

Beyond Large Language Models: A Mental World Model Framework for Mental Modeling, Simulation, Evolution, and Interpretability (arxiv.org)

OwlSight: A Panoramic Monitor for Chain-of-Thought Reasoning (arxiv.org)

GEMS: Molecular Structure Elucidation as Learned Search over Isomer Graphs (arxiv.org)

From Trajectories to Differentiated Skills: Cognitive Differentiation for Self-Improving Agents (arxiv.org)

RL-Guided Temporal Localization for Dual-Channel Retrieval in Long-Horizon Agent Memory (arxiv.org)

Darwin-Claw: Environment-Adaptive Experience Consolidation for Recursively Self-Evolving Agents (arxiv.org)

SkillBrowser: Selective Skill Exposure for Scalable Skill-Using Agents (arxiv.org)

Trapped in Your Story: Evaluating Narrative-Induced Sycophancy in Large Language Models (arxiv.org)

LocalIntent: Outcome-Specific Structured Generation for Backend-Compatible Smart Building Automation with Local LLMs (arxiv.org)

S-GAP: Stealthy Graph Attack Primitives for Graph Large Language Models (arxiv.org)

GatewayBench: Unified Capability and Security Evaluation Platform for MCP Gateways (arxiv.org)

ScoutRoute: Efficient Mixture-of-Experts Routing Using Internal Language Model Signals (arxiv.org)

Generative Policy Improvement from Observational Data (arxiv.org)

Monitor-Validity Failures: LLM Safety Evidence Paths at Evolving Framework Boundaries (arxiv.org)

OliO: ODE-based Linear Transition Operator for Self-Supervised Time Series Forecasting (arxiv.org)

Robust Function Calling for LLM Agents Under Interface Drift (arxiv.org)

Expressivity of IQP Feature-Map Quantum Kernels via Reproducing Kernel Hilbert Spaces (arxiv.org)

VStress: Correlation-Aware Auditing and Adaptive Budget Allocation for Repeated Verifiers (arxiv.org)

Pretraining Transformers with Quantized Softmax in Attention (arxiv.org)

Group-Graph Policy Optimization for Long-Horizon Agentic Reinforcement Learning (arxiv.org)

Learning reveals invisible structure in low-rank RNNs (arxiv.org)

AgentOnto: Ontology-Guided LLM Agents for Enterprise Data Analytics (arxiv.org)

From Memory Hits to Answerable Evidence: Query-Conditioned Memory Compilation for Long-Term Agent Memory (arxiv.org)

Self-Reflective Chain-of-Thought (SRCoT): Evidence-Based Text Generation via a Closed-Loop Retrieve--Reflect--Revise Pipeline on Edge LLMs (arxiv.org)

SimOPD: Relational Knowledge Transfer for On-Policy Distillation (arxiv.org)

Optimist Deep Research: An Iterative, State-Centric Agent for Autonomous Literature Research (arxiv.org)

ToMStories: Interpreting Theory-of-Mind in Synthetic Narratives (arxiv.org)

CoTA++: Understanding the Repeat Curse in dMLLMs from an Information Flow Perspective (arxiv.org)

Annotrace: A Workbench for Step-Level Annotation of Agent Traces (arxiv.org)

Reasoning Models Can Control Their Chains of Thought When They Tell Themselves To (arxiv.org)

Beyond Reward Scoring: Online Multi-Agent Preference Learning for Multimodal Response Generation (arxiv.org)

Evo-B&B: Behavior-Aware LLM-Guided Joint Evolution of Branching and Node Selection Policies for MILP (arxiv.org)

World Action Critic (WAC) Model: Joint Foresight and Criticism for Policy Learning (arxiv.org)

Compression Precedes, Content Decides: Emergent Misalignment in Reward Models (arxiv.org)

When Can You Trust an Agent's Own Verifier? Stress-Testing Counterfactual Rationale Verification Under Search Pressure and Oracle Access (arxiv.org)

StoBasket: Joint Upsell and Cross-Sell Optimization for Contribution Margin (arxiv.org)

Deterministic Bounds on Accuracy and Stability of Metric-Space Regression and Classification (arxiv.org)