Blog
No more endless PDFs. Discover the core value of the latest top-tier research in one article.
Tunneling from an oscillating initial state in quantum mechanics
Relit-LiVE: Relight Video by Jointly Learning Environment Video
Positivity of the gravitational path integral implies the axionic weak gravity conjecture
Gravitational Lensing of Gravitational Waves from Astrophysical Sources: Theory, Detection, and Applications
Towards Steering without Sacrifice: Principled Training of Steering Vectors for Prompt-only Interventions
Active Learning for Conditional Generative Compressed Sensing
MemReranker: Reasoning-Aware Reranking for Agent Memory Retrieval
Teaching Thinking Models to Reason with Tools: A Full-Pipeline Recipe for Tool-Integrated Reasoning
OA-WAM: Object-Addressable World Action Model for Robust Robot Manipulation
MUSE: Resolving Manifold Misalignment in Visual Tokenization via Topological Orthogonality
The multi-fractal nature of pedestrian arrival times
MedHorizon: Towards Long-context Medical Video Understanding in the Wild
EMO: Pretraining Mixture of Experts for Emergent Modularity
Superintelligent Retrieval Agent: The Next Frontier of Information Retrieval
Recursive Agent Optimization
WavCube: Unifying Speech Representation for Understanding and Generation via Semantic-Acoustic Joint Modeling
Beyond Semantic Similarity: Rethinking Retrieval for Agentic Search via Direct Corpus Interaction
LatentRAG: Latent Reasoning and Retrieval for Efficient Agentic RAG
Optimizer-Model Consistency: Full Finetuning with the Same Optimizer as Pretraining Forgets Less
Retrieval from Within: An Intrinsic Capability of Attention-Based Models
Attribution-Guided Continual Learning for Large Language Models
TIDE: Every Layer Knows the Token Beneath the Context
Instrumental Choices: Measuring the Propensity of LLM Agents to Pursue Instrumental Behaviors
StraTA: Incentivizing Agentic Reinforcement Learning with Strategic Trajectory Abstraction
Nonsense Helps: Prompt Space Perturbation Broadens Reasoning Exploration
RecGPT-Mobile: On-Device Large Language Models for User Intent Understanding in Taobao Feed Recommendation
Sharpness-Aware Pretraining Mitigates Catastrophic Forgetting
Efficient Pre-Training with Token Superposition
PrefixGuard: From LLM-Agent Traces to Online Failure-Warning Monitors
VideoRouter: Query-Adaptive Dual Routing for Efficient Long-Video Understanding