AI Research Papers — Page 76

SANA-WM: Efficient Minute-Scale World Modeling with Hybrid Linear Diffusion Transformer

OpenDeepThink: Parallel Reasoning via Bradley--Terry Aggregation

MetaBackdoor: Exploiting Positional Encoding as a Backdoor Attack Surface in LLMs

Evidential Reasoning Advances Interpretable Real-World Disease Screening

Text Knows What, Tables Know When: Clinical Timeline Reconstruction via Retrieval-Augmented Multimodal Alignment

Does Synthetic Layered Design Data Benefit Layered Design Decomposition?

Position: Behavioural Assurance Cannot Verify the Safety Claims Governance Now Demands

Hand-in-the-Loop: Improving Dexterous VLA via Seamless Interventional Correction

MeMo: Memory as a Model

Self-Distilled Agentic Reinforcement Learning

RoSHAP: A Distributional Framework and Robust Metric for Stable Feature Attribution

Pelican-Unified 1.0: A Unified Embodied Intelligence Model for Understanding, Reasoning, Imagination and Action

Widening the Gap: Exploiting LLM Quantization via Outlier Injection

Causal Forcing++: Scalable Few-Step Autoregressive Diffusion Distillation for Real-Time Interactive Video Generation

Forgetting That Sticks: Quantization-Permanent Unlearning via Circuit Attribution

Training ML Models with Predictable Failures

Causal Foundation Models with Continuous Treatments

APWA: A Distributed Architecture for Parallelizable Agentic Workflows

Natural Synthesis: Outperforming Reactive Synthesis Tools with Large Reasoning Models

MemEye: A Visual-Centric Evaluation Framework for Multimodal Agent Memory

Understanding How International Students in the U.S. Are Using Conversational AI to Support Cross-Cultural Adaptation

CoCo-InEKF: State Estimation with Learned Contact Covariances in Dynamic, Contact-Rich Scenarios

CLOVER: Closed-Loop Value Estimation & Ranking for End-to-End Autonomous Driving Planning

Problem This paper addresses the gap in the training and evaluation methodologies for end-to-end autonomous driving planners, particularly the mismatch between training on a single logged trajectory and evaluating using…

Talk is (Not) Cheap: A Taxonomy and Benchmark Coverage Audit for LLM Attacks

DriveCtrl: Conditioned Sim-to-Real Driving Video Generation

Learning from Language Feedback via Variational Policy Distillation

Proposal and study of statistical features for string similarity computation and classification

Why Neighborhoods Matter: Traversal Context and Provenance in Agentic GraphRAG

Logging Policy Design for Off-Policy Evaluation

From Text to Voice: A Reproducible and Verifiable Framework for Evaluating Tool Calling LLM Agents