AI.info

Latest Research — Page 18

Browse Latest Research on AI.info.

• EngramEdit: Decoupled Knowledge Updates in LLMs through Conditional Memory
• Advancing Brain Tumor Segmentation via Attention-based 3D U-Net Architecture and Digital Image Processing
• AnyAD: Unified Any-Modality Anomaly Detection in Incomplete Multi-Sequence MRI
• Gromov-Wasserstein Distillation for Inductive Multi-View Embedding
• Towards Scalable Oversight via Partitioned Human Supervision
• PAT3D: Physics-Augmented Text-to-3D Scene Generation
• On-Policy Parameter Update Direction Underlies Generalization in LLM Post-Training
• When Personalization Tricks Detectors: The Feature-Inversion Trap in Machine-Generated Text Detection
• Certain Head, Uncertain Tail: Expert-Sample for Test-Time Scaling in Fine-Grained MoE
• Incoherence in Goal-Conditioned Autoregressive Models
• Interpretable Causal Discovery via Causal-Effect Constraints
• Line-based Event Preprocessing: Towards Low-Energy Neuromorphic Computer Vision
• Adaptive Testing for LLM Evaluation: A Psychometric Alternative to Static Benchmarks
• BhashaBench V1: A Comprehensive Benchmark for the Quadrant of Indic Domains
• MemWeaver: A Hierarchical Memory from Textual Interactive Behaviors for Personalized Generation
• XGait: A Multi-Modality Wireless Sensing Dataset for Indoor Human Tracking and Identification
• MTR-DuplexBench: Towards a Comprehensive Evaluation of Multi-Round Conversations for Full-Duplex Speech Language Models
• Architect-Ant: Editable Automatic Furnishing of Architectural Floor Plans
• Vision-Language-Action Models for Autonomous Driving: Past, Present, and Future
• mmWalk: Towards Multi-modal Multi-view Walking Assistance
• Diff4Splat: Controllable 4D Scene Generation with Latent Dynamic Reconstruction Models
• Sparse Additive Model Pruning for Order-Based Causal Structure Learning
• Memento 3: Model-Based Recursive Self-Improvement through Reflective Rulebooks
• SPICE: Submodular Penalized Information-Conflict Selection for Efficient Large Language Model Training
• SANE: Automated Scorer-Specific Metamorphic Testing for LLM Judges
• Prioritize the Process, Not Just the Outcome: Rewarding Latent Thought Trajectories Improves Reasoning in Looped Language Models
• HiFOR: Hierarchical Foresight with Action-Guided World Grounding for Vision-Language-Action Model
• MoniRefer: A Real-world Large-scale Multi-modal Dataset based on Roadside Infrastructure for 3D Visual Grounding
• Distributed Subliminal Learning: Replacing Model Updates with Random-Carrier Outputs
• One Bias After Another: Mechanistic Reward Shaping and Persistent Biases in Language Reward Models
• Transferring Causal Effects using Proxies
• The Smart Buildings Challenge: Learning Foundation Models for Real-World Optimization at Scale
• On Evaluating LLM Alignment by Evaluating LLMs as Judges
• MyoMechanix: Biomechanically-Grounded Compositional Skilled Activity Understanding and Coaching
• Partition Tree: Conditional Density Estimation over General Outcome Spaces
• Deep Search in Real-World Web Environments
• OverSeer: Multi-Agent World Models Beyond What Agents See
• Random Attention: Rethinking KV Cache Eviction for Efficient Reasoning
• Do Vision-Language Models Understand Visual Persuasiveness? A Diagnosis via Visual Persuasive Factors
• FlexRank: Nested Low-Rank Knowledge Decomposition for Adaptive Model Deployment