AI.info

Latest Research — Page 85

Browse Latest Research on AI.info.

• Logical Phase Transitions: Understanding Collapse in LLM Logical Reasoning
• NOTAM-Evolve: A Knowledge-Guided Self-Evolving Optimization Framework with LLMs for NOTAM Interpretation
• Deep Incomplete Multi-View Clustering via Hierarchical Imputation and Alignment
• From Inpainting to Editing: Unlocking Robust Mask-Free Visual Dubbing via Generative Bootstrapping
• Beyond Fixed Tasks: Unsupervised Environment Design for Task-Level Pairs
• VLM-Loc: Localization in Point Cloud Maps via Vision-Language Models
• Joint Speech and Text Training for LLM-Based End-to-End Spoken Dialogue State Tracking
• Connecting the Dots: Training-Free Visual Grounding via Agentic Reasoning
• Color3D: Controllable and Consistent 3D Colorization with Personalized Colorizer
• E-Scores for (In)Correctness Assessment of Generative Model Outputs
• Augmenting Biological Fitness Prediction Benchmarks with Landscapes Features from GraphFLA
• A Standardized Benchmark for Multilabel Antimicrobial Peptide Classification
• Align Once, Benefit Multilingually: Enforcing Multilingual Consistency for LLM Safety Alignment
• Secu-Table: a Comprehensive security table dataset for evaluating semantic table interpretation systems
• Design Techniques for LLM-Powered Interactive Storytelling: A Case Study of the Dramamancer System
• ReSplat: Learning Recurrent Gaussian Splatting
• Scaling and Transferability of Annealing Strategies in Large Language Model Training
• Zero-Shot Textual Explanations via Translating Decision-Critical Features
• Time Series Forecasting via Direct Per-Step Probability Distribution Modeling
• TabICLv2: A better, faster, scalable, and open tabular foundation model
• OTI: A Model-free and Visually Interpretable Measure of Image Attackability
• The Horcrux: Mechanistically Interpretable Task Decomposition for Detecting and Mitigating Reward Hacking in Embodied AI Systems
• SimBench: Benchmarking the Ability of Large Language Models to Simulate Human Behaviors
• Demo: Guide-RAG: Evidence-Driven Corpus Curation for Retrieval-Augmented Generation in Long COVID
• Reliable and Responsible Foundation Models: A Comprehensive Survey
• Diffusion Knows Transparency: Repurposing Video Diffusion for Transparent Object Depth and Normal Estimation
• InfiniHuman: Infinite 3D Human Creation with Precise Control
• Knowledge-guided generative surrogate modeling for high-dimensional design optimization under scarce data
• Large Language Models Develop Novel Social Biases Through Adaptive Exploration
• LGDC: Latent Graph Diffusion via Spectrum-Preserving Coarsening
• REFA: Real-time Egocentric Facial Animations for Virtual Reality
• Co-Annotator: Expert-Distilled ViT and VLM for Visual and Documentation Guidance in Age-Related Macular Degeneration
• CENIC: Convex Error-controlled Numerical Integration for Contact
• Hallucination Begins Where Saliency Drops
• A Random Matrix Theory Perspective on the Consistency of Diffusion Models
• The Collaboration Gap: Exploration and Benchmarking of Open-World Agentic Cooperation
• MAFA: A Multi-Agent Framework for Enterprise-Scale Annotation with Configurable Task Adaptation
• Composing Concepts from Images and Videos via Concept-prompt Binding
• Training One Model to Master Cross-Level Agentic Actions via Reinforcement Learning
• EchoMind: An Interrelated Multi-level Benchmark for Evaluating Empathetic Speech Language Models