-
Notifications
You must be signed in to change notification settings - Fork 41
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[20210718] Weekly AI ArXiv 만담 #17
Comments
Papers
News
|
Paper
|
DeepMind’s AI for protein structure is coming to the masses (Nature)
Evaluating Large Language Models Trained on Code
IS GPT-3 OVERHYPED? Austin commits to $1.5B for DOD’s Joint AI Center over next 5 years
중동 지역은 차세대 AI 허브가 될 수 있을 것인가 인공지능 시대 사기의 아이콘 된 AI 투자 중국 인공지능 특허량 세계 1위
|
What Classifiers Don't Know What They Don't? Facebook에서 Model Uncertainty 및 Calibration에 대해 다양한 방법을 비교한 실험 논문입니다. ImageNot Dataset이라는 ImageNet을 ResNet18의 feature를 기준으로 OOD로 나눈 데이터셋을 구별하고 in-distribution과 out-of-distribution data에 대해 최근 제시된 다양한 방법들을 비교합니다. Passive attention in artificial neural networks predicts human visual selectivity Deep Mind에서 CAM 등 soft-attention 방법들과 실제 인간의 시각 등 확인하는 영역이 일치하는지 실험적으로 확인하는 심리학과와의 협업 논문입니다. Deep Learning Theory 보다는 인간의 attention과 Deep Learning의 attention이 일치하는지 확인하는 논문입니다. Per-Pixel Classification is Not All You Need for Semantic Segmentation Facebook 에서 pixel-level classification이 아닌 mask prediction을 기존 architecture에 바로 사용할 수 있도록 구현한 새로운 방법론입니다. Segmentation 을 하시는 분들께 도우이 될 것 같아 공유드립니다. Facebook에 의하면 현재 ADE20K에서 SOTA라고 합니다. Revisiting the Calibration of Modern Neural Networks Fully Sharded Data Parallel: faster AI training with fewer GPUs |
Regularization is all you Need: Simple Neural Nets can Excel on Tabular 좀 다른 의견의 논문인데 요것도 흥미있어 보입니다. |
AutoGluon: https://auto.gluon.ai/stable/index.html |
AutoGluon Tabular 논문: https://arxiv.org/abs/2003.06505 |
|
Conservative Objective Models for Effective Offline Model-Based Optimization ICML 21
|
LabML AI 링크 공유해드립니다. https://papers.labml.ai/papers/daily |
AI News
AI ArXiv
Blender Bot 2.0
Industry and Academic Research in Computer Vision
ViTGAN: Training GANs with Vision Transformers
ERNIE 3.0: Large-scale Knowledge Enhanced Pre-training for Language Understanding and Generation
FLEX: Unifying Evaluation for Few-Shot NLP
Randomized ReLU Activation for Uncertainty Estimation of Deep Neural Networks
Spanish Language Models
AFHQ-v2 dataset: https://github.com/clovaai/stargan-v2
The text was updated successfully, but these errors were encountered: