Blog
No more endless PDFs. Discover the core value of the latest top-tier research in one article.
AnchorRefine: Synergy-Manipulation Based on Trajectory Anchor and Residual Refinement for Vision-Language-Action Models
MM-Hand: A 21-DOF Multi-modal Modular Dexterous Robotic Hand with Remote Actuation
AnyLift: Scaling Motion Reconstruction from Internet Videos via 2D Diffusion
MemSearch-o1: Empowering Large Language Models with Reasoning-Aligned Memory Growth in Agentic Search
DINOv3 Beats Specialized Detectors: A Simple Foundation Model Baseline for Image Forensics
A Mechanistic Analysis of Sim-and-Real Co-Training in Generative Robot Policies
HiVLA: A Visual-Grounded-Centric Hierarchical Embodied Manipulation System
From Seeing to Simulating: Generative High-Fidelity Simulation with Digital Cousins for Generalizable Robot Learning and Evaluation
LiteResearcher: A Scalable Agentic RL Training Framework for Deep Research Agent
CogGen: A Cognitively Inspired Recursive Framework for Deep Research Report Generation
CAPO: Counterfactual Credit Assignment in Sequential Cooperative Teams
Classical and quantum evolution of inflationary fluctuations
Speculative Decoding for Autoregressive Video Generation
AnyRecon: Arbitrary-View 3D Reconstruction with Video Diffusion Model
CoInteract: Physically-Consistent Human-Object Interaction Video Synthesis via Spatially-Structured Co-Generation
LoopCTR: Unlocking the Loop Scaling Power for Click-Through Rate Prediction
UniMesh: Unifying 3D Mesh Understanding and Generation
XEmbodied: A Foundation Model with Enhanced Geometric and Physical Cues for Large-Scale Embodied Environments
It's all in your head -- fine-tuning arguments do not require aleatoric uncertainty
IR-Flow: Bridging Discriminative and Generative Image Restoration via Rectified Flow
RAD-2: Scaling Reinforcement Learning in a Generator-Discriminator Framework
Generative Drifting for Conditional Medical Image Generation
SpanVLA: Efficient Action Bridging and Learning from Negative-Recovery Samples for Vision-Language-Action Model
GRAFT: Geometric Refinement and Fitting Transformer for Human Scene Reconstruction
TrafficClaw: Generalizable Urban Traffic Control via Unified Physical Environment Modeling
From Gaia to GaiaNIR: II. A new view of the Milky Way bar
StepPO: Step-Aligned Policy Optimization for Agentic Reinforcement Learning
SpaceDex: Generalizable Dexterous Grasping in Tiered Workspaces
Beyond Rating: A Comprehensive Evaluation and Benchmark for AI Reviews
OmniHuman: A Large-scale Dataset and Benchmark for Human-Centric Video Generation