AI News

When Do Learned Blocks Preserve Causal Meaning? Block Identifiability and Constructive Causal Abstractions (arxiv.org)

Claim Framing as a Driver of LLM Sycophancy (arxiv.org)

LoOPSD: On-Policy Self-Distillation with Looped Teachers (arxiv.org)

Looped Reasoning Finishes Earlier Than You Think (arxiv.org)

A Few More Tokens Ahead: Reinforcing Parallelism in Diffusion LLMs (arxiv.org)

Encrypted Skills Can Become New Security Exploits If AI Agents Know How to Decipher (arxiv.org)

JevGPT: Thinking Fast and Slow in Autonomous Research (arxiv.org)

Which Optimizer Leaks Most? Optimizer Implicit Bias as a Training-Data Privacy Controller (arxiv.org)

Teaching MLP More Graph Information: A Three-stage Multitask Knowledge Distillation Framework (arxiv.org)

AffectWorld: Benchmarking Emotion-Driven Future World Prediction (arxiv.org)

SimInter: Evidence-Grounded User Simulation with Interview-Based Profile Enrichment for Conversational Recommender Systems (arxiv.org)

ShoeFit: A New Dataset and Dual-image-stream DiT Framework for Virtual Footwear Try-On (arxiv.org)

Lightweight but Linguistically Fair? Quantisation Robustness across Bavarian, German, and English (arxiv.org)

Let the Mind Conceive, Let the Hand Achieve: Following Multimodal Instructions for Long-Horizon Robotic Manipulation (arxiv.org)

Urbex: Agentic Spatial Grounding for City-Scale 3D Scenes (arxiv.org)

PRISM: Calibrating Black-Box Judges with Counterfactual Probes (arxiv.org)

Graph-free Retrieval with Graph-Modified Embeddings (arxiv.org)

Not All Drift Is a Mechanism: Tracing Prompt Adaptation in LLMs (arxiv.org)

Do Factuality Methods Make LLMs More Factual? Reproducibility and Fair Comparison (arxiv.org)

Resource-Efficient 3D Shape and Orientation Estimation of Road Users using Roadside RGB Camera and 3D Radar (arxiv.org)

Agaram 2.5B: The First Tamil-First Language Model Trained from Scratch with Tool Calling. A Playbook for Under-Represented Languages (arxiv.org)

FedGraph: Defending Federated Large Language Model Fine-Tuning Against Backdoor Attacks via Graph-Based Aggregation (arxiv.org)

How Well Do LLMs Perform in Peer Review? Evidence from Finance and Accounting (arxiv.org)

STAMP as a topology-aware molecular language for fragment-level learning and editing (arxiv.org)

Learning Perturbation-Specific Representations through Conditional Energy Prior for Cellular Response Prediction (arxiv.org)

MemoryVLN: From Ego-Aware Memory Representation to Cross-Context Memory Reuse in Vision-Language Navigation (arxiv.org)

Task-evoked eye movements as candidate digital biomarkers of depression severity in adolescents and young adults (arxiv.org)

Better Capturing Interactions Between Products in~Retail: Revisited Negative Sampling for~Basket Choice Modeling (arxiv.org)

EgoOmniWorld: Unifying Egocentric Understanding via Ego-Action-World Interactions (arxiv.org)

A Global View of How LLMs Value Human Time (arxiv.org)

The Geometry of Language Model Logits: Generic Structure and Learned Departures (arxiv.org)

LLM2R: From Large Language Models to 3D Reassemblers (arxiv.org)

APTree: Extended Behavior Trees for Adaptive and Scalable Automated Robotic Task Planning in Construction (arxiv.org)

On Symmetry for Characterizing Agent Heterogeneity in Multi-Agent Systems (arxiv.org)

The Many Morals of AI: The cautionary bias of North American news narratives in global perspective (arxiv.org)

When More Objectives Make Identification Easier: The Role of Preferences and Feedback (arxiv.org)

Falsification-proof non-market allocation mechanisms (arxiv.org)

GeoMER: A Joint Euclidean-Hyperbolic Geometry Framework for Multimodal Emotion Recognition (arxiv.org)

ACCRUE: Continual Off-Policy Reinforcement Learning for Robot Foundation Models (arxiv.org)

PhysicsSliders: Learning Physical Controls for Audio Foley Generation (arxiv.org)

Rethinking Projector Training in Multimodal LLMs (arxiv.org)

Model-Based Online Decision Making via Generative Trajectory Planning (arxiv.org)

DPOison-VL: Clean Preference Poisoning in Vision-Language Direct Preference Optimization (arxiv.org)

Jev-ToM: Evaluating Theory of Mind in Typed Decision Models (arxiv.org)

Nudge the Judge: Uncovering Program Decision Hijacking in Real-World Jev Applications (arxiv.org)

Does self-distillation improve reasoning? (arxiv.org)

Knowledge Graphs as Files: An Agent Harness Where Tool Creation and Tool Use Co-Evolve (arxiv.org)

Multi-agent Undercover Gaming: Hallucination Removal via Counterfactual Test for Multimodal Reasoning (arxiv.org)

Alignment Faking - the Train -> Deploy Asymmetry: Through a Game-Theoretic Lens with Bayesian-Stackelberg Equilibria (arxiv.org)

AlignMerge - Alignment-Preserving Large Language Model Merging via Fisher-Guided Geometric Constraints (arxiv.org)