Blog
No more endless PDFs. Discover the core value of the latest top-tier research in one article.
Streaming Autoregressive Video Generation via Diagonal Distillation
PostTrainBench: Can LLM Agents Automate LLM Post-Training?
AtomicVLA: Unlocking the Potential of Atomic Skill Learning in Robots
Reinforced Generation of Combinatorial Structures: Ramsey Numbers
When to Lock Attention: Training-Free KV Control in Video Diffusion
TiPToP: A Modular Open-Vocabulary Planning System for Robotic Manipulation
Robust Regularized Policy Iteration under Transition Uncertainty
DISPLAY: Directable Human-Object Interaction Video Generation via Sparse Motion Guidance and Multi-Task Auxiliary
Reading, Not Thinking: Understanding and Bridging the Modality Gap When Text Becomes Pixels in Multimodal LLMs
Reject, Resample, Repeat: Understanding Parallel Reasoning in Language Model Inference
Interactive World Simulator for Robot Policy Training and Evaluation
CARE-Edit: Condition-Aware Routing of Experts for Contextual Image Editing
TaSR-RAG: Taxonomy-guided Structured Reasoning for Retrieval-Augmented Generation
Boosting MLLM Spatial Reasoning with Geometrically Referenced 3D Scene Representations
Towards a Neural Debugger for Python
See, Plan, Rewind: Progress-Aware Vision-Language-Action Models for Robust Robotic Manipulation
Computer Vision-Based Vehicle Allotment System using Perspective Mapping
Has quantum advantage been achieved?
ImageEdit-R1: Boosting Multi-Agent Image Editing via Reinforcement Learning
Variational Flow Maps: Make Some Noise for One-Step Conditional Generation
LiveWorld: Simulating Out-of-Sight Dynamics in Generative Video World Models
HY-WU (Part I): An Extensible Functional Neural Memory Framework and An Instantiation in Text-Guided Image Editing
From Data Statistics to Feature Geometry: How Correlations Shape Superposition
$Δ$VLA: Prior-Guided Vision-Language-Action Models via World Knowledge Variation
ConfCtrl: Enabling Precise Camera Control in Video Diffusion via Confidence-Aware Interpolation
TeamHOI: Learning a Unified Policy for Cooperative Human-Object Interactions with Any Team Size
Reviving ConvNeXt for Efficient Convolutional Diffusion Models
Think Before You Lie: How Reasoning Improves Honesty
Speed3R: Sparse Feed-forward 3D Reconstruction Models
Chain of Event-Centric Causal Thought for Physically Plausible Video Generation