Blog
No more endless PDFs. Discover the core value of the latest top-tier research in one article.
Region-Constrained Group Relative Policy Optimization for Flow-Based Image Editing
Sharp description of local minima in the loss landscape of high-dimensional two-layer ReLU neural networks
Wireless Communication Enhanced Value Decomposition for Multi-Agent Reinforcement Learning
AdaCubic: An Adaptive Cubic Regularization Optimizer for Deep Learning
LatentFlowSR: High-Fidelity Audio Super-Resolution via Noise-Robust Latent Flow Matching
ActiveGlasses: Learning Manipulation with Active Vision from Ego-centric Human Demonstration
Seeing is Believing: Robust Vision-Guided Cross-Modal Prompt Learning under Label Noise
The missing ultra-faint satellites of the Milky Way
E-3DPSM: A State Machine for Event-Based Egocentric 3D Human Pose Estimation
Task-Aware LLM Routing with Multi-Level Task-Profile-Guided Data Synthesis for Cold-Start Scenarios
Toward Hardware-Agnostic Quadrupedal World Models via Morphology Conditioning
HiFloat4 Format for Language Model Pre-training on Ascend NPUs
EGLOCE: Training-Free Energy-Guided Latent Optimization for Concept Erasure
The Cosmic Web and Its Filaments: Neutrino Mass from Topology and Persistent Homology
Decoupling Vector Data and Index Storage for Space Efficiency
Nonlocal Games Revisited: A Representation-Theoretic Path from Bell Locality to Quantum Pseudo-Telepathy
Effective strings and particles interacting in 3D: the Ising model
The Illusory Precision of TTV Masses: Hidden Solutions Behind Kepler-9's Tight Mass Ratio
Structure-Aware Fine-Grained Gaussian Splatting for Expressive Avatar Reconstruction
SPPO: Sequence-Level PPO for Long-Horizon Reasoning Tasks
Enhancing Conversational TTS with Cascaded Prompting and ICL-Based Online Reinforcement Learning
The four-loop non-singlet splitting functions in QCD
Physically Grounded 3D Generative Reconstruction under Hand Occlusion using Proprioception and Multi-Contact Touch
Arbitration Failure, Not Perceptual Blindness: How Vision-Language Models Resolve Visual-Linguistic Conflicts
MeshOn: Intersection-Free Mesh-to-Mesh Composition
Quantum Uncertainty and Entropy
Skill-Conditioned Visual Geolocation for Vision-Language
Space- vs Time-dependence in taming the infrared instability of projectable Ho\v rava Gravity
Repurposing 3D Generative Model for Autoregressive Layout Generation
FineCog-Nav: Integrating Fine-grained Cognitive Modules for Zero-shot Multimodal UAV Navigation