Blog
No more endless PDFs. Discover the core value of the latest top-tier research in one article.
InCoder-32B: Code Foundation Model for Industrial Scenarios
The PokeAgent Challenge: Competitive and Long-Context Learning at Scale
Nemotron-Cascade 2: Post-Training LLMs with Cascade RL and Multi-Domain On-Policy Distillation
SWE-QA-Pro: A Representative Benchmark and Scalable Training Recipe for Repository-Level Code Understanding
Manifold-Matching Autoencoders
LoST: Level of Semantics Tokenization for 3D Shapes
Generation Models Know Space: Unleashing Implicit 3D Priors for Scene Understanding
Versatile Editing of Video Content, Actions, and Dynamics without Training
AgentFactory: A Self-Evolving Framework Through Executable Subagent Accumulation and Reuse
AR-CoPO: Align Autoregressive Video Generation with Contrastive Policy Optimization
Generative Control as Optimization: Time Unconditional Flow Matching for Adaptive and Robust Robotic Control
Symphony: A Cognitively-Inspired Multi-Agent System for Long-Video Understanding
V-Dreamer: Automating Robotic Simulation and Trajectory Synthesis via Video Generation Priors
A brief introduction to Poisson geometry
Procedural Generation of Algorithm Discovery Tasks in Machine Learning
Memento-Skills: Let Agents Design Agents
CUBE: A Standard for Unifying Agent Benchmarks
Kestrel: Grounding Self-Refinement for LVLM Hallucination Mitigation
Reasoning over mathematical objects: on-policy reward modeling and test time aggregation
Promotion and rowmotion in rational Catalan combinatorics
VideoAtlas: Navigating Long-Form Video in Logarithmic Compute
DreamPlan: Efficient Reinforcement Fine-Tuning of Vision-Language Planners via Video World Models
Video Understanding: From Geometry and Semantics to Unified Models
EVA: Aligning Video World Models with Executable Robot Actions via Inverse Dynamics Rewards
PanoVGGT: Feed-Forward 3D Reconstruction from Panoramic Imagery
GUI-CEval: A Hierarchical and Comprehensive Chinese Benchmark for Mobile GUI Agents
Efficient Reasoning on the Edge
VisBrowse-Bench: Benchmarking Visual-Native Search for Multimodal Browsing Agents
ProbeFlow: Training-Free Adaptive Flow Matching for Vision-Language-Action Models
S-VAM: Shortcut Video-Action Model by Self-Distilling Geometric and Semantic Foresight