AI.info
Latest Research — Page 60
Browse Latest Research on AI.info.
- Cataract-LMM Large-Scale Multi-Source Multi-Task Benchmark for Deep Learning in Surgical Video Analysis
- Agent-in-the-Loop: A Data Flywheel for Continuous Improvement in LLM-based Customer Support
- Prefill-Free Cross-Family KV Cache Transfer for Heterogeneous Multi-Agent LLMs
- Spatio-Temporal Context Learning with Temporal Difference Convolution for Moving Infrared Small Target Detection
- Replacing Training with Memory: Listwise Selection for Text-to-SQL
- Safin-1: Safety from Within through Memory-Native State Evolution
- Towards Generative Location Awareness for Disaster Response: A Probabilistic Cross-view Geolocalization Approach
- Can Test-Time Memory Evolution Outperform Test-Time Training for Scientific Discovery?
- Navigating the Alignment-Calibration Trade-off: A Pareto-Superior Frontier via Model Merging
- ACT as Human: Multimodal Large Language Model Data Annotation with Critical Thinking
- Difficulty-Aware Label-Guided Denoising for Monocular 3D Object Detection
- Internalizing Agent Experience into Diffusion Model Weights via On-Policy Context Distillation
- G2P: Gaussian-to-Point Attribute Alignment for Boundary-Aware 3D Segmentation
- Mobile-4DGS: Unified Static-Dynamic Real-time Mobile Gaussian Splatting
- DarkQA: Benchmarking Vision-Language Models on Visual-Primitive Question Answering in Low-Light Indoor Scenes
- Reward Modeling for Reinforcement Learning-Based LLM Reasoning: Design, Challenges, and Evaluation
- Generalized CycleGRPO for Fine-Grained Multimodal Understanding and Grounding
- How Sampling Affects the Detectability of Machine-written texts: A Comprehensive Study
- Each Reward to Its Relevant Tokens: Reward-Specific Token-Level Credit Assignment for Multi-Reward RL
- Actions Have Consequences: Detecting Outcome Performativity using Intervention Testing
- Learning When to Trust via Selective Context Preference Optimization
- Latent Collaboration in Multi-Agent Systems
- Understanding Multimodality in Generative Behavioral Cloning
- VisJudge-Bench: Aesthetics and Quality Assessment of Visualizations
- TagaVLM: Topology-Aware Global Action Reasoning for Vision-Language Navigation
- GRACE: Generation-aware latent compression for efficient video generation
- Backbone-Informed Adaptation after EEG Pretraining: A Capacity-Controlled Study
- AutoSynth: Automated Workflow Optimization for High-Quality Synthetic Dataset Generation via Monte Carlo Tree Search
- WFR-FM: Simulation-Free Dynamic Unbalanced Optimal Transport
- InfiniteWeb: Scalable Web Environment Synthesis for GUI Agent Training
- Objective vs. Search: Decomposing What Makes a Good Tokeniser
- Adapting Without Gradients: Affine Statistics Transport and What Its Certificate Can Tell You
- Beyond Runtime Enforcement: Shield Synthesis as Defensibility Analysis for Adversarial Networks
- Self-OPD: On-Policy Distillation for Flow Matching Models without Teacher
- World Action Modeling with Progressive Visual Planning
- Robot-Assisted Group Tours for Blind People
- Measurement-First Auditing of Agentic Leaderboards: Contamination Susceptibility, Matched-Control Re-evaluation, and Scorer Validation
- Rethinking Video Generation Model for the Embodied World
- Training LLM Judges from Language Feedback via Position-Selective Self-Distillation
- Inter-Agent Trust Models: A Comparative Study of Brief, Claim, Proof, Stake, Reputation and Constraint in Agentic Web Protocol Design-A2A, AP2, ERC-8004, and Beyond