Blog
No more endless PDFs. Discover the core value of the latest top-tier research in one article.
RoboCasa365: A Large-Scale Simulation Framework for Training and Benchmarking Generalist Robots
Single-minus graviton tree amplitudes are nonzero
(Quantum) reference frames, relational observables, gauge reduction and physical interpretation
The Stellar Mass Function for Nine Massive Galaxy Clusters in the Local Universe
Force-Aware Residual DAgger via Trajectory Editing for Precision Insertion with Impedance Control
MoD-DPO: Towards Mitigating Cross-modal Hallucinations in Omni LLMs using Modality Decoupled Preference Optimization
What Does Flow Matching Bring To TD Learning?
Any2Any: Unified Arbitrary Modality Translation for Remote Sensing
MoECLIP: Patch-Specialized Experts for Zero-shot Anomaly Detection
Beyond the Grid: Layout-Informed Multi-Vector Retrieval with Parsed Visual Document Representations
ACE-Merging: Data-Free Model Merging with Adaptive Covariance Estimation
SkillVLA: Tackling Combinatorial Diversity in Dual-Arm Manipulation via Skill Reuse
IDProxy: Cold-Start CTR Prediction for Ads and Recommendation at Xiaohongshu with Multimodal LLMs
$V_1$: Unifying Generation and Self-Verification for Parallel Reasoners
Inherited Goal Drift: Contextual Pressure Can Undermine Agentic Goals
RealWonder: Real-Time Physical Action-Conditioned Video Generation
Analysis of the Riemann Zeta Function via Recursive Taylor Expansions
MOOSE-Star: Unlocking Tractable Training for Scientific Discovery by Breaking the Complexity Barrier
A Rubric-Supervised Critic from Sparse Real-World Outcomes
RoboPocket: Improve Robot Policies Instantly with Your Phone
FlashAttention-4: Algorithm and Kernel Pipelining Co-Design for Asymmetric Hardware Scaling
DARE: Aligning LLM Agents with the R Statistical Ecosystem via Distribution-Aware Retrieval
Mixture of Universal Experts: Scaling Virtual Width via Depth-Width Transformation
Timer-S1: A Billion-Scale Time Series Foundation Model with Serial Scaling
Locality-Attending Vision Transformer
Towards Multimodal Lifelong Understanding: A Dataset and Agentic Baseline
KARL: Knowledge Agents via Reinforcement Learning
Solving an Open Problem in Theoretical Physics using AI-Assisted Discovery
Planning in 8 Tokens: A Compact Discrete Tokenizer for Latent World Model
DeepScan: A Training-Free Framework for Visually Grounded Reasoning in Large Vision-Language Models