AI.info

Latest Research — Page 6

Browse Latest Research on AI.info.

• Rethinking Relative Policy Differences Transfer of On-Policy Distillation
• Mind-to-Face: Neural-Driven Photorealistic Avatar Synthesis via EEG Decoding
• InformedXRD: Reproducible Benchmarks and Physics-Informed Evaluation for Powder Diffraction Symmetry Classification
• MotorSense: A Video-EMG Dataset of Motoric Representations for Action Understanding
• Improved Masked Image Generation with Knowledge-Augmented Token Representations
• Revisiting [CLS] and Patch Token Interaction in Vision Transformers
• Controlling Long-Horizon Behavior in Language Model Agents with Explicit State Dynamics
• QuCo-RAG: Quantifying Uncertainty from the Pre-training Corpus for Dynamic Retrieval-Augmented Generation
• Adaptive Confidence Regularization for Multimodal Failure Detection
• Understanding the Effects of Distractors on Reasoning Vision-Language Models
• Know your Trajectory -- Trustworthy Reinforcement Learning deployment through Importance-Based Trajectory Analysis
• PixelUMM: Encoder-Free Unified Image and Video Understanding and Generation
• The False Promise of Zero-Shot Super-Resolution in Machine-Learned Operators
• MultiHateLoc: Towards Temporal Localisation of Multimodal Hate Content in Online Videos
• Look and Tell: A Dataset for Multimodal Grounding Across Egocentric and Exocentric Views
• Gender Bias in Vision-Language In-Context Learning
• Bias Is a Subspace, Not a Coordinate: A Geometric Rethinking of Post-hoc Debiasing in Vision-Language Models
• GraspLDP: Towards Generalizable Grasping Policy via Latent Diffusion
• JUICE: Unified Multimodal Modeling in Pretrained Representation Spaces
• The Impact of Quantization on Large Reasoning Model Reinforcement Learning
• Med-CMR: A Fine-Grained Benchmark Integrating Visual Evidence and Clinical Logic for Medical Complex Multimodal Reasoning
• Enhancing Generalization of Depth Estimation Foundation Model via Weakly-Supervised Adaptation with Regularization
• Predict Training Data Quality via Its Geometry in Metric Space
• Multimodal Bandits: Regret Lower Bounds and Optimal Algorithms
• CompanionHarm: A Multi-Turn Benchmark for Detecting Harms in Real-World AI Companion Conversations
• From Detection to Association: Learning Discriminative Object Embeddings for Multi-Object Tracking
• From SERPs to Agents: A Platform for Comparative Studies of Information Interaction
• How Stable Are LLM-Generated Proofs Under Program Refactoring? A Stress Test for Verifiable Code Generation
• Agentic AI for Scalable and Robust Optical Systems Control
• CoFEH: LLM-driven Feature Engineering Empowered by Collaborative Bayesian Hyperparameter Optimization
• A2RL V\textsubscript{max}: The A2RL autonomous racing dataset for long-range, high-speed perception and multi-vehicle interaction
• Relic: From Multi-Agent Collaboration to Persistent Organizational Capability
• DeepAnalyze: Agentic Large Language Models for Autonomous Data Science
• MERGETUNE: Continued Fine-Tuning of Vision-Language Models
• Multi-Crit: Benchmarking Multimodal Judges on Pluralistic Criteria-Following
• EvoSafeHarness: Evolving Model- and Domain-Specific Harnesses for Securing Agents
• Enrich and Detect: Video Temporal Grounding with Multimodal LLMs
• Anatomy Contextualized Adaption of CT Foundation Models
• New Lenses on Learning Design: Course Types and Indicators from Large-Scale Moodle Data
• What Git Does Not Record: Provenance Legibility in Two AI-Assisted Research Repositories