AI.info

Latest Research — Page 72

Browse Latest Research on AI.info.

• Flattening Every Memory Peak in Long-Context Mixture-of-Experts Training
• LadderEdit: Edit-Level Residual Compression for Memory-Efficient Lifelong Editing of LLMs
• Vernata: Self-Supervised Learning of LiDAR Point Representations
• Do Compact SSL Backbones Matter for Audio Deepfake Detection? A Controlled Study with RAPTOR
• SynthWorlds: Controlled Parallel Worlds for Disentangling Reasoning and Knowledge in Language Models
• From Parameter to Representation: A Closed-Form Approach for Controllable Model Merging
• HSKBenchmark: Modeling and Benchmarking Chinese Second Language Acquisition in Large Language Models through Curriculum Tuning
• CtrlCache: Accelerating Interactive Video World Models with Control-Aware Caching
• Efficient Edge-Compatible CNN for Speckle-Based Material Recognition in Laser Cutting Systems
• HuGe: Scaling Humanoid Generalist Vision-Language-Action Models with Egocentric Experience
• LiTo: Surface Light Field Tokenization
• Statistical Inference for Explainable Boosting Machines
• Coherence-Driven Belief Formation and Population Dynamics of Contagion in LLM Agents
• Benford's Law as a Distributional Prior for Post-Training Quantization of Large Language Models
• ThinkTank-ME: A Multi-Expert Framework for Middle East Event Forecasting
• An RL View of OPD: Least Square Policy Distillation for Sample-Efficient LLM Reasoning
• Hyperparameter Transfer Enables Consistent Gains of Matrix-Preconditioned Optimizers Across Scales
• Online Action-Stacking Improves Reinforcement Learning Performance for Air Traffic Control
• Pluggable Pruning with Contiguous Layer Distillation for Diffusion Transformers
• PulseLM: A Foundation Dataset and Benchmark for PPG-Text Learning
• Training Specialist Models without Reasoning Trajectories for Domain Expert Distillation
• On the Token Collapse Phenomenon in Textual Graph Soft Prompting
• Why Do Multilingual Reasoning Gaps Emerge in Reasoning Language Models?
• Does Visual Grounding Enhance the Understanding of Embodied Knowledge in Large Language Models?
• K-Sort Eval: Efficient Preference Evaluation for Visual Generation via Corrected VLM-as-a-Judge
• A Deep Multi-Modal Method for Patient Wound Healing Assessment
• Rethinking Composed Image Retrieval Evaluation: A Fine-Grained Benchmark from Image Editing
• Rudder: Steering Prefetching in Distributed GNN Training using LLM Agents
• Training High-Level Schedulers with Execution-Feedback Reinforcement Learning for Long-Horizon GUI Automation
• An Open and Reproducible Deep Research Agent for Long-Form Question Answering
• On the Information Processing of One-Dimensional Wasserstein Distances with Finite Samples
• End-to-end example-based sim-to-real RL policy transfer based on neural stylisation with application to robotic cutting
• TinyScientist: An Interactive, Extensible, and Controllable Framework for Building Research Agents
• Progress-Aware Distillation for Mitigating Stagnation in Small Language Model Agents
• UrbanAI 2025 Challenge: Linear vs Transformer Models for Long-Horizon Exogenous Temperature Forecasting
• PhysSFI-Net: Physics-informed Geometric Learning of Skeletal and Facial Interactions for Orthognathic Surgical Outcome Prediction
• ConSensus: Multi-Agent Collaboration for Multimodal Sensing
• Scalable Random Wavelet Features: Efficient Non-Stationary Kernel Approximation with Convergence Guarantees
• Quantum-Enhanced Generative Models for Rare Event Prediction
• STEAM: A Semantic-Level Knowledge Editing Framework for Large Language Models