This repo contains my paper reading notes on deep learning. (mainly on NLP)
Check the notes at https://bc-li.github.io/paperreading.
| Title | Field | Time | Report link | Time I started | Status |
|---|---|---|---|---|---|
| [ICCV 2015] Learning Deconvolution Network for Semantic Segmentation | VISION | 2015 | https://bc-li.github.io/paper/deconvnet | 2021/5/17 | Done |
| [NeurIPS 2017] Attention Is All You Need | NLP | 2017 | https://bc-li.github.io/paper/transformer | 2021/12/11 | Done |
| [NAACL 2019] BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding | NLP | 2018 | https://bc-li.github.io/paper/bert | 2021/12/15 | Done |
| [NeurIPS 2014] Sequence to Sequence Learning with Neural Networks | NLP | 2014 | https://bc-li.github.io/paper/seq2seq | 2022/1/21 | Done |
| [ICLR 2018] Non-Autoregressive Neural Machine Translation | NLP | 2018 | https://bc-li.github.io/paper/nonauto | 2022/1/24 | Done |
| [ICLR 2019] Parameter-Efficient Transfer Learning for NLP | NLP | 2019 | https://bc-li.github.io/paper/petl | 2022/2/2 | Done |
| [ICLR 2018] Unsupervised Neural Machine Translation | NLP | 2018 | https://bc-li.github.io/paper/unsupervised-NMT | 2022/2/4 | Done |
| Classic CNN structures (LeNet to DenseNet) | VISION | 1998-2017 | https://bc-li.github.io/paper/cnn | 2022/2/25 | Done |
| MobileNets series [V1 to V3] | VISION | 2017 | https://bc-li.github.io/paper/mobilenets | 2022/3/10 | Done |
| SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size | VISION | 2016 | https://bc-li.github.io/paper/SqueezeNet | 2022/3/10 | Dropped |
| [NeurIPS 2019] Levenshtein Transformer | NLP | 2019 | https://bc-li.github.io/paper/lt | 2022/2/15 | 70% |
| [NeurIPS 2021] Multimodal Few-Shot Learning with Frozen Language Models | MultiModal | 2021 | https://bc-li.github.io/paper/frozenmodel | 2022/4/6 | 50% |
| [ICCV 2021] Pano-AVQA: Grounded Audio-Visual Question Answering on 360◦ Videos | MultiModal | 2021 | https://bc-li.github.io/paper/pano-avqa | 2022/4/15 | Pending |
| [CVPR 2022] Learning to Answer Questions in Dynamic Audio-Visual Scenarios | MultiModal | 2022 | https://bc-li.github.io/paper/music-avqa | 2022/4/15 | 80% |
| [CVPR 2021] Taming Transformers for High-Resolution Image Synthesis | MultiModal | 2021 | https://bc-li.github.io/paper/vqgan | 2022/8/4 | Done |
| [NeurIPS 2020] Denoising Diffusion Probabilistic Models | ML | 2020 | https://bc-li.github.io/paper/ddpm | 2022/10/3 | Pending |
| [ICLR 2022] LoRA: Low-Rank Adaptation of Large Language Models | ML | 2022 | https://bc-li.github.io/paper/lora | 2023/4/21 | Pending |
| Title | Field | Time |
|---|---|---|
| [SCTS 2020] Pre-trained Models for Natural Language Processing: A Survey | NLP | 2020 |
| [EMNLP 2020] A Simple and Effective Model for Answering Multi-span Questions | NLP | 2020 |
| [NAACL 2022] OPERA: Operation-Pivoted Discrete Reasoning over Text | NLP | 2022 |
| [NAACL 2019] Drop: A reading comprehension benchmark requiring discrete reasoning over paragraphs | NLP | 2019 |
| [EMNLP 2020] Question directed graph attention network for numerical reasoning over text | NLP | 2020 |
| [NeurIPS 2022] Diffusion-LM Improves Controllable Text Generation | NLP | 2022 |
写 blog 的时候如未特殊说明则为从约为零基础开始。在 blog post 中我会把我为了理解文中一些比较 specific 的概念找到的相对容易理解的原出处贴到文中，方便查阅，且不再重复阐述。
写 blog 的时候会参阅的资料和感觉比较好的 UP 主。