Blog
告别冗长 PDF，一文读懂最新顶会与核心期刊的创新价值。
AlphaFlowTSE: One-Step Generative Target Speaker Extraction via Conditional AlphaFlow
IH-Challenge: A Training Dataset to Improve Instruction Hierarchy on Frontier LLMs
Adaptive Loops and Memory in Transformers: Think Harder or Know More?
ActiveUltraFeedback: Efficient Preference Data Generation using Active Learning
Differentiable Geometric Indexing for End-to-End Generative Retrieval
Motion Forcing: A Decoupled Framework for Robust Video Generation in Motion Dynamics
Electromagnetic Signatures of Supermassive Binary Black Holes. I. Thermal Synchrotron, Self-Lensing Flares, and Jet Precession
Quantum entanglement provides a competitive advantage in adversarial games
Is this Idea Novel? An Automated Benchmark for Judgment of Research Ideas
One-Eval: An Agentic System for Automated and Traceable LLM Evaluation
HTMuon: Improving Muon via Heavy-Tailed Spectral Correction
DynamicVGGT: Learning Dynamic Point Maps for 4D Scene Reconstruction in Autonomous Driving
FireRedASR2S: A State-of-the-Art Industrial-Grade All-in-One Automatic Speech Recognition System
MUGEN: Evaluating and Improving Multi-audio Understanding of Large Audio-Language Models
Neural Field Thermal Tomography: A Differentiable Physics Framework for Non-Destructive Evaluation
When Fine-Tuning Fails and when it Generalises: Role of Data Diversity and Mixed Training in LLM-based TTS
All-Loop Renormalization and the Phase of the de Sitter Wavefunction
From Prior to Pro: Efficient Skill Mastery via Distribution Contractive RL Finetuning
TA-Mem: Tool-Augmented Autonomous Memory Retrieval for LLM in Long-Term Conversational QA
S2D: Sparse to Dense Lifting for 3D Reconstruction with Minimal Inputs
AgentOS: From Application Silos to a Natural Language-Driven Data Ecosystem
GLM-OCR Technical Report
Aligning Large Language Models with Searcher Preferences
LegONet: Plug-and-Play Structure-Preserving Neural Operator Blocks for Compositional PDE Learning
Hybrid Self-evolving Structured Memory for GUI Agents
Quality over Quantity: Demonstration Curation via Influence Functions for Data-Centric Robot Learning
OmniGuide: Universal Guidance Fields for Enhancing Generalist Robot Policies
Beyond Scalars: Evaluating and Understanding LLM Reasoning via Geometric Progress and Stability
A$^2$-Edit: Precise Reference-Guided Image Editing of Arbitrary Objects and Ambiguous Masks
UniStitch: Unifying Semantic and Geometric Features for Image Stitching