7월에 San Diego에서 열리는 NeurIPS 2025에 다음 2편의 논문을 spotlight로 발표하게 되었습니다.
"Rethinking Entropy in Test-Time Adaptation: The Missing Piece from Energy Duality"
"Shortcut Features as Top Eigenfunctions of NTK: A Linear Neural Network Case and More"
KDST는 딥러닝을 포함한 데이터와 지능에 관련된 여러 주제에 대해서 연구하는 팀입니다.
KDST
KDST
tag cloud
- 강화학습
- pruning
- gan
- MLP-Mixer
- robustness
- bearing fault detection
- Reinforcement Learning
- autoencoder
- point-based image editing #dragdiffusion #dragondiffusion
- DQN
- DegAug
- domain shift
- NaturalInversion
- Diffusion Model
- Singulairty
- Manifold Regularized Dynamic Network Pruning
- 2022AAAI
- Data-Free Image Synthesis
- representation leanring
- Dynamic Pruning
- 2022-AAAI
- bearing fault diagnosis
- anomaly detection
- cnn
- Out of Distribution
- AAAI 2021
- ICCV
- Knowledge Distillation
- Sharpness-Aware Minimization
- domain generalization
