CHAI: A CHatbot AI for Task-Oriented Dialogue with Offline Reinforcement Learning
— AK (@ak92501) April 19, 2022
abs: https://t.co/HirMgPmieI pic.twitter.com/L88ZBDZJZn
CHAI: A CHatbot AI for Task-Oriented Dialogue with Offline Reinforcement Learning
— AK (@ak92501) April 19, 2022
abs: https://t.co/HirMgPmieI pic.twitter.com/L88ZBDZJZn
An Extendable, Efficient and Effective Transformer-based Object Detector
— AK (@ak92501) April 19, 2022
abs: https://t.co/3D2aSqmSkr
github: https://t.co/tNopT866Jc pic.twitter.com/cZN3Sbooob
LayoutLMv3: Pre-training for Document AI with Unified Text and Image Masking
— AK (@ak92501) April 19, 2022
abs: https://t.co/wuzHfvfDHQ
github: https://t.co/dms3SfhNQo pic.twitter.com/PDd8Xfp1K9
mGPT: Few-Shot Learners Go Multilingual
— AK (@ak92501) April 19, 2022
abs: https://t.co/9uxHVoqRXO
introduces two autoregressive GPT-like models with 1.3 billion and 13 billion parameters trained on 60 languages
from 25 language families using Wikipedia and Colossal Clean Crawled Corpus pic.twitter.com/gDGX6qjv8A
Improving Passage Retrieval with Zero-Shot Question Generation
— AK (@ak92501) April 18, 2022
abs: https://t.co/rZjAJoqfzb pic.twitter.com/UNpy6ERqdG
Training Compute-Optimal Large Language Models — “We find that current large language models are significantly undertrained” 👀 https://t.co/UEjVWhBFyq
— Sebastian Raschka (@rasbt) April 16, 2022
Exhaustive Survey of Rickrolling in Academic Literature
— AK (@ak92501) April 15, 2022
abs: https://t.co/pm0XBhwf0L
video: https://t.co/Mcli4ZYVUO pic.twitter.com/x4nkElT4BB
Neighborhood Attention Transformer
— AK (@ak92501) April 15, 2022
abs: https://t.co/8IHy2LfbjE
Experimental results on NAT are competitive; NAT-Tiny reaches 83.2% top-1 accuracy on ImageNet with only 4.3 GFLOPs and 28M parameters, 51.4% mAP on MS-COCO and 48.4% mIoU on ADE20k pic.twitter.com/M7rUHOM4Tr
Masked Siamese Networks for Label-Efficient Learning
— AK (@ak92501) April 15, 2022
abs: https://t.co/dYXpFnTm3Y
github: https://t.co/MHm8z6lBWr
on ImageNet-1K, with only 5,000 annotated images, base MSN model achieves 72.4% top-1 accuracy, and with 1% of ImageNet-1K labels, achieves 75.7% top-1
accuracy pic.twitter.com/wXhSeUtNc5
"Know Your Limits: Uncertainty Estimation with ReLU Classifiers Fails at Reliable OOD Detection" -- an interesting paper on this subject where the authors have a theoretical explanation that ReLU and Softmax are (partly) to blame: https://t.co/STbOjj9YJx pic.twitter.com/pEoztyQRHw
— Sebastian Raschka (@rasbt) April 14, 2022
InCoder: A Generative Model for Code Infilling and Synthesis
— AK (@ak92501) April 14, 2022
abs: https://t.co/qAbrJzgVkw
project page: https://t.co/Sp87l2oGix pic.twitter.com/U0iNz40ZWq
Few-shot Learning with Noisy Labels
— AK (@ak92501) April 13, 2022
abs: https://t.co/wGnBAoCH8D
results show that TraNFS is on-par with leading FSL methods on clean support sets, yet outperforms them, by far, in the presence of label noise pic.twitter.com/HdIFPzpQKM