arXiv V1: Constitutional AI: Harmlessness from AI Feedback
Read MorearXiv V1: FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awareness
arXiv V1: Beyond 3DMM: Learning to Capture High-fidelity 3D Face Shape
arXiv V1: Training Compute-Optimal Large Language Models
arXiv V1: Training language models to follow instructions with human feedback
arXiv V1: Scaling Language Models: Methods, Analysis & Insights from Training Gopher 120 pages
arXiv V2: LORA: LOW-RANK ADAPTATION OF LARGE LANGUAGE MODELS Github -
arXiv V1: Linear Transformers Are Secretly Fast Weight Programmers
Published as a conference paper at ICLR 2021 arXiv V1: AN IMAGE IS WORTH 16X16 WORDS: TRANSFORMERS FOR IMAGE RECOGNITION AT SCALE
arXiv V2: Beyond Self-attention: External Attention using Two Linear Layers for Visual Tasks
Published as a conference paper at ICCV 2021 ICCV Open Access Version: Swin Transformer: Hierarchical Vision Transformer using Shi...
arXiv V2: On Layer Normalization in the Transformer Architecture
arXiv V2: BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding BERT, which stands for Bidirectional Enc...
arXiv V1: BERT for Joint Intent Classification and Slot Filling
Published as a conference paper at ICLR 2019 arXiv V2: LARGE SCALE GAN TRAINING FOR HIGH FIDELITY NATURAL IMAGE SYNTHESIS
arXiv V1: Glow: Generative Flow with Invertible 1×1 Convolutions Github
Published as a conference paper at ICLR 2017 arXiv V6: BI-DIRECTIONAL ATTENTION FLOW FOR MACHINE COMPREHENSION 抽取式问答 首次提出了双向的at...
OpenAI: Improving Language Understanding by Generative Pre-Training 抽取式问答 首次提出了双向的attention流应用在MRC任务， 在SQuAD数据上取得sota的结果，后面的许多论文都是...
arXiv V1: Universal Language Model Fine-tuning for Text Classification