% Define an abbreviation (acronym) % Acronyms for indonesian thesis \newacronym{ml}{ML}{\textit{machine learning}} \newacronym{stft}{STFT}{\textit{short-time} Fourier \textit{transform}} \newacronym{ai}{AI}{\textit{artificial intelligence}} \newacronym{dl}{DL}{\textit{deep learning}} \newacronym{nn}{NN}{\textit{neural network}} \newacronym{fft}{FFT}{\textit{fast} Fourier \textit{transform}} \newacronym{svm}{SVM}{\textit{support vector machine}} \newacronym{cnn}{CNN}{\textit{convolutional neural network}} \newacronym{rnn}{RNN}{\textit{recurrent neural network}} \newacronym{vbi}{VBI}{\textit{vibration-based inspection}} \newacronym{shm}{SHM}{\textit{structural health monitoring}} \newacronym{fea}{FEA}{\textit{finite element analysis}} \newacronym{1d-cnn}{1-D CNN}{\textit{one-dimensional convolutional neural network}} \newacronym{pca}{PCA}{\textit{principal component analysis}} \newacronym{rbf}{RBF}{\textit{radial basis function}} \newacronym{cm}{CM}{\textit{confusion matrix}} \newacronym{tsne}{t-SNE}{\textit{t-Distributed Stochastic Neighbor Embedding}} \newacronym{pacmap}{PaCMAP}{\textit{Pairwise Controlled Manifold Approximation Projection}} \newacronym{ar}{AR}{\textit{autoregressive}} \newacronym{re}{RE}{\textit{residual error}} \newacronym{ae}{AE}{\textit{autoencoder}} \newacronym{lr}{LR}{\textit{logistic regression}} \newacronym{ann}{ANN}{\textit{artificial neural network}} \newacronym{wt}{WT}{\textit{wavelet transform}} \newacronym{oc-svm}{OC-SVM}{\textit{one-class support vector machine}} \newacronym{lstm}{LSTM}{\textit{long short-term memory}} \newacronym{ga}{GA}{\textit{genetic algorithm}} \newacronym{ht}{HT}{Hilbert \textit{transform}} \newacronym{vmd}{VMD}{\textit{variational mode decomposition}} \newacronym{qugs}{QUGS}{\textit{Qatar University Grandstand Simulator}} \newacronym{r-cnn}{R-CNN}{\textit{region-based convolutional neural network}} \newacronym{pso}{PSO}{\textit{particle swarm optimization}} % ADDED - Common AI/ML acronyms with proper names \newacronym{hmm}{HMM}{\textit{hidden} Markov \textit{model}} \newacronym{gru}{GRU}{\textit{gated recurrent unit}} \newacronym{gan}{GAN}{\textit{generative adversarial network}} \newacronym{vae}{VAE}{\textit{variational autoencoder}} \newacronym{adam}{Adam}{Adam} % optimizer named after "adaptive moment estimation" \newacronym{sgd}{SGD}{\textit{stochastic gradient descent}} \newacronym{relu}{ReLU}{\textit{rectified linear unit}} \newacronym{elu}{ELU}{\textit{exponential linear unit}} \newacronym{selu}{SELU}{\textit{scaled exponential linear unit}} \newacronym{gelu}{GELU}{Gaussian \textit{error linear unit}} \newacronym{bert}{BERT}{\textit{Bidirectional Encoder Representations from Transformers}} \newacronym{gpt}{GPT}{\textit{Generative Pre-trained Transformer}} \newacronym{resnet}{ResNet}{\textit{Residual Network}} \newacronym{yolo}{YOLO}{\textit{You Only Look Once}} \newacronym{fcn}{FCN}{\textit{fully convolutional network}} \newacronym{rlhf}{RLHF}{\textit{reinforcement learning from human feedback}} \newacronym{ppo}{PPO}{\textit{proximal policy optimization}} \newacronym{dqn}{DQN}{\textit{deep Q-network}} \newacronym{ddpg}{DDPG}{\textit{deep deterministic policy gradient}} \newacronym{a3c}{A3C}{\textit{asynchronous advantage actor-critic}} \newacronym{sac}{SAC}{\textit{soft actor-critic}} \newacronym{td}{TD}{\textit{temporal difference}} \newacronym{mdp}{MDP}{Markov \textit{decision process}} \newacronym{pomdp}{POMDP}{\textit{partially observable} Markov \textit{decision process}} \newacronym{mcmc}{MCMC}{Markov \textit{chain} Monte Carlo} \newacronym{em}{EM}{\textit{expectation-maximization}} \newacronym{map}{MAP}{\textit{maximum a posteriori}} \newacronym{mle}{MLE}{\textit{maximum likelihood estimation}} \newacronym{kl}{KL}{Kullback-Leibler} \newacronym{js}{JS}{Jensen-Shannon} \newacronym{knn}{k-NN}{\textit{k-nearest neighbors}} \newacronym{dbscan}{DBSCAN}{\textit{density-based spatial clustering of applications with noise}} \newacronym{kmeans}{k-means}{\textit{k-means}} \newacronym{gmm}{GMM}{Gaussian \textit{mixture model}} \newacronym{gp}{GP}{Gaussian \textit{process}} \newacronym{gpr}{GPR}{Gaussian \textit{process regression}} \newacronym{bo}{BO}{Bayesian \textit{optimization}} \newacronym{tpe}{TPE}{\textit{tree-structured} Parzen \textit{estimator}} \newacronym{smbo}{SMBO}{\textit{sequential model-based optimization}} \newacronym{nas}{NAS}{\textit{neural architecture search}} \newacronym{automl}{AutoML}{\textit{automated machine learning}} \newacronym{xgboost}{XGBoost}{\textit{eXtreme Gradient Boosting}} \newacronym{gbdt}{GBDT}{\textit{gradient boosting decision tree}} \newacronym{rf}{RF}{\textit{random forest}} \newacronym{dt}{DT}{\textit{decision tree}} \newacronym{cart}{CART}{\textit{classification and regression tree}} \newacronym{id3}{ID3}{\textit{Iterative Dichotomiser 3}} \newacronym{c45}{C4.5}{C4.5} \newacronym{adaboost}{AdaBoost}{\textit{adaptive boosting}} \newacronym{lasso}{LASSO}{\textit{least absolute shrinkage and selection operator}} \newacronym{ridge}{Ridge}{Ridge \textit{regression}} \newacronym{enet}{ElasticNet}{ElasticNet} \newacronym{svr}{SVR}{\textit{support vector regression}} \newacronym{kpca}{KPCA}{\textit{kernel principal component analysis}} \newacronym{lda}{LDA}{\textit{linear discriminant analysis}} \newacronym{ica}{ICA}{\textit{independent component analysis}} \newacronym{nmf}{NMF}{\textit{non-negative matrix factorization}} \newacronym{umap}{UMAP}{\textit{Uniform Manifold Approximation and Projection}} \newacronym{isomap}{Isomap}{\textit{Isometric Mapping}} \newacronym{lle}{LLE}{\textit{locally linear embedding}} \newacronym{mds}{MDS}{\textit{multidimensional scaling}} \newacronym{som}{SOM}{\textit{self-organizing map}} \newacronym{bpnn}{BPNN}{\textit{backpropagation neural network}} \newacronym{elm}{ELM}{\textit{extreme learning machine}} \newacronym{sae}{SAE}{\textit{stacked autoencoder}} \newacronym{dae}{DAE}{\textit{denoising autoencoder}} \newacronym{vq-vae}{VQ-VAE}{\textit{vector quantized variational autoencoder}} \newacronym{wgan}{WGAN}{Wasserstein GAN} \newacronym{dcgan}{DCGAN}{\textit{deep convolutional generative adversarial network}} \newacronym{cgan}{cGAN}{\textit{conditional generative adversarial network}} \newacronym{cyclegan}{CycleGAN}{CycleGAN} \newacronym{pix2pix}{Pix2Pix}{Pix2Pix} \newacronym{stylegan}{StyleGAN}{StyleGAN} \newacronym{clip}{CLIP}{\textit{Contrastive Language-Image Pre-training}} \newacronym{vit}{ViT}{\textit{Vision Transformer}} \newacronym{swin}{Swin}{\textit{Shifted Window Transformer}} \newacronym{detr}{DETR}{\textit{DEtection TRansformer}} \newacronym{rcnn}{R-CNN}{\textit{region-based convolutional neural network}} \newacronym{fast-rcnn}{Fast R-CNN}{Fast R-CNN} \newacronym{faster-rcnn}{Faster R-CNN}{Faster R-CNN} \newacronym{mask-rcnn}{Mask R-CNN}{Mask R-CNN} \newacronym{fpn}{FPN}{\textit{feature pyramid network}} \newacronym{ssd}{SSD}{\textit{single shot detector}} \newacronym{seresnet}{SE-ResNet}{\textit{Squeeze-and-Excitation ResNet}} \newacronym{nlp}{NLP}{\textit{natural language processing}} \newacronym{nlu}{NLU}{\textit{natural language understanding}} \newacronym{nlg}{NLG}{\textit{natural language generation}} \newacronym{seq2seq}{Seq2Seq}{\textit{sequence-to-sequence}} \newacronym{bleu}{BLEU}{\textit{Bilingual Evaluation Understudy}} \newacronym{rouge}{ROUGE}{\textit{Recall-Oriented Understudy for Gisting Evaluation}} \newacronym{meteor}{METEOR}{\textit{Metric for Evaluation of Translation with Explicit ORdering}} \newacronym{cider}{CIDEr}{\textit{Consensus-based Image Description Evaluation}} \newacronym{wer}{WER}{\textit{word error rate}} \newacronym{cer}{CER}{\textit{character error rate}} \newacronym{perplexity}{PPL}{\textit{perplexity}} \newacronym{bpe}{BPE}{\textit{byte pair encoding}} \newacronym{wordpiece}{WordPiece}{WordPiece} \newacronym{sentencepiece}{SentencePiece}{SentencePiece} \newacronym{tfidf}{TF-IDF}{\textit{term frequency-inverse document frequency}} \newacronym{bow}{BoW}{\textit{bag of words}} \newacronym{cbow}{CBOW}{\textit{continuous bag of words}} \newacronym{skipgram}{Skip-gram}{Skip-gram} \newacronym{word2vec}{Word2Vec}{Word2Vec} \newacronym{glove}{GloVe}{\textit{Global Vectors for Word Representation}} \newacronym{fasttext}{FastText}{FastText} \newacronym{elmo}{ELMo}{\textit{Embeddings from Language Models}} \newacronym{ulmfit}{ULMFiT}{\textit{Universal Language Model Fine-tuning}} \newacronym{t5}{T5}{\textit{Text-to-Text Transfer Transformer}} \newacronym{bart}{BART}{\textit{Bidirectional and Auto-Regressive Transformers}} \newacronym{electra}{ELECTRA}{\textit{Efficiently Learning an Encoder that Classifies Token Replacements Accurately}} \newacronym{roberta}{RoBERTa}{\textit{Robustly Optimized BERT Pretraining Approach}} \newacronym{ernie}{ERNIE}{\textit{Enhanced Representation through kNowledge IntEgration}} \newacronym{flash-attention}{FlashAttention}{FlashAttention} \newacronym{mha}{MHA}{\textit{multi-head attention}} \newacronym{mqa}{MQA}{\textit{multi-query attention}} \newacronym{gqa}{GQA}{\textit{grouped-query attention}} \newacronym{kv-cache}{KV-cache}{\textit{key-value cache}} \newacronym{rag}{RAG}{\textit{retrieval-augmented generation}} \newacronym{dpr}{DPR}{\textit{dense passage retrieval}} \newacronym{faiss}{FAISS}{\textit{Facebook AI Similarity Search}} \newacronym{annoy}{Annoy}{\textit{Approximate Nearest Neighbors Oh Yeah}} \newacronym{hnsw}{HNSW}{\textit{Hierarchical Navigable Small World}} \newacronym{ivf}{IVF}{\textit{inverted file index}} \newacronym{pq}{PQ}{\textit{product quantization}} \newacronym{lsh}{LSH}{\textit{locality-sensitive hashing}} \newacronym{mips}{MIPS}{\textit{maximum inner product search}} \newacronym{rl}{RL}{\textit{reinforcement learning}} \newacronym{drl}{DRL}{\textit{deep reinforcement learning}} \newacronym{irl}{IRL}{\textit{inverse reinforcement learning}} \newacronym{gail}{GAIL}{\textit{generative adversarial imitation learning}} \newacronym{il}{IL}{\textit{imitation learning}} \newacronym{bc}{BC}{\textit{behavioral cloning}} \newacronym{marl}{MARL}{\textit{multi-agent reinforcement learning}} \newacronym{maddpg}{MADDPG}{\textit{multi-agent deep deterministic policy gradient}} \newacronym{vdn}{VDN}{\textit{value decomposition network}} \newacronym{coma}{COMA}{\textit{counterfactual multi-agent policy gradient}} \newacronym{mappo}{MAPPO}{\textit{multi-agent proximal policy optimization}} \newacronym{trpo}{TRPO}{\textit{trust region policy optimization}} \newacronym{td3}{TD3}{\textit{twin delayed deep deterministic policy gradient}} \newacronym{her}{HER}{\textit{hindsight experience replay}} \newacronym{per}{PER}{\textit{prioritized experience replay}} \newacronym{c51}{C51}{\textit{categorical 51}} \newacronym{qr-dqn}{QR-DQN}{\textit{quantile regression DQN}} \newacronym{iqn}{IQN}{\textit{implicit quantile network}} \newacronym{icm}{ICM}{\textit{intrinsic curiosity module}} \newacronym{rnd}{RND}{\textit{random network distillation}} \newacronym{ngu}{NGU}{\textit{Never Give Up}} \newacronym{curl}{CURL}{\textit{Contrastive Unsupervised Representations for Reinforcement Learning}} \newacronym{mcts}{MCTS}{Monte Carlo \textit{tree search}} \newacronym{uct}{UCT}{\textit{upper confidence bound applied to trees}} \newacronym{puct}{PUCT}{\textit{predictor + upper confidence bound applied to trees}} \newacronym{ucb}{UCB}{\textit{upper confidence bound}} \newacronym{ts}{TS}{Thompson \textit{sampling}} \newacronym{mab}{MAB}{\textit{multi-armed bandit}} \newacronym{cmab}{CMAB}{\textit{contextual multi-armed bandit}} \newacronym{exp3}{EXP3}{\textit{exponential-weight algorithm for exploration and exploitation}} \newacronym{linucb}{LinUCB}{\textit{linear upper confidence bound}} \newacronym{scaffold}{SCAFFOLD}{\textit{Stochastic Controlled Averaging for Federated Learning}} \newacronym{fedbn}{FedBN}{\textit{Federated learning with Batch Normalization}} \newacronym{moon}{MOON}{\textit{Model-Contrastive Federated Learning}} \newacronym{feddf}{FedDF}{\textit{Federated Learning via Knowledge Distillation with Dynamic Regularization}} \newacronym{fl}{FL}{\textit{federated learning}} \newacronym{dp}{DP}{\textit{differential privacy}} \newacronym{ldp}{LDP}{\textit{local differential privacy}} \newacronym{pate}{PATE}{\textit{Private Aggregation of Teacher Ensembles}} \newacronym{smpc}{SMPC}{\textit{secure multi-party computation}} \newacronym{he}{HE}{\textit{homomorphic encryption}} \newacronym{phe}{PHE}{\textit{partial homomorphic encryption}} \newacronym{fhe}{FHE}{\textit{fully homomorphic encryption}} \newacronym{mpc}{MPC}{\textit{multi-party computation}} \newacronym{tee}{TEE}{\textit{trusted execution environment}} \newacronym{sgx}{SGX}{\textit{Software Guard Extensions}} \newacronym{zkp}{ZKP}{\textit{zero-knowledge proof}} \newacronym{snark}{SNARK}{\textit{Succinct Non-interactive ARgument of Knowledge}} \newacronym{stark}{STARK}{\textit{Scalable Transparent ARgument of Knowledge}} \newacronym{zksnark}{zkSNARK}{\textit{zero-knowledge Succinct Non-interactive ARgument of Knowledge}} \newacronym{zkstark}{zkSTARK}{\textit{zero-knowledge Scalable Transparent ARgument of Knowledge}} \newacronym{mlops}{MLOps}{\textit{Machine Learning Operations}} \newacronym{cicd}{CI/CD}{\textit{continuous integration/continuous deployment}} \newacronym{dag}{DAG}{\textit{directed acyclic graph}} \newacronym{etl}{ETL}{\textit{extract, transform, load}} \newacronym{elt}{ELT}{\textit{extract, load, transform}} \newacronym{oltp}{OLTP}{\textit{online transaction processing}} \newacronym{olap}{OLAP}{\textit{online analytical processing}} \newacronym{dwh}{DWH}{\textit{data warehouse}} \newacronym{datalake}{Data Lake}{\textit{data lake}} \newacronym{lakehouse}{Lakehouse}{\textit{lakehouse}} \newacronym{edw}{EDW}{\textit{enterprise data warehouse}} \newacronym{dim}{DIM}{\textit{dimension table}} \newacronym{fact}{FACT}{\textit{fact table}} \newacronym{scd}{SCD}{\textit{slowly changing dimension}} \newacronym{star-schema}{Star Schema}{\textit{star schema}} \newacronym{snowflake-schema}{Snowflake Schema}{\textit{snowflake schema}} \newacronym{eer}{EER}{\textit{equal error rate}} \newacronym{far}{FAR}{\textit{false acceptance rate}} \newacronym{frr}{FRR}{\textit{false rejection rate}} \newacronym{fpr}{FPR}{\textit{false positive rate}} \newacronym{tpr}{TPR}{\textit{true positive rate}} \newacronym{fnr}{FNR}{\textit{false negative rate}} \newacronym{tnr}{TNR}{\textit{true negative rate}} \newacronym{ppv}{PPV}{\textit{positive predictive value}} \newacronym{npv}{NPV}{\textit{negative predictive value}} \newacronym{auroc}{AUROC}{\textit{area under the receiver operating characteristic curve}} \newacronym{auprc}{AUPRC}{\textit{area under the precision-recall curve}} \newacronym{roc}{ROC}{\textit{receiver operating characteristic}} \newacronym{pr-curve}{PR Curve}{\textit{precision-recall curve}} \newacronym{f1}{F1}{\textit{F1 score}} \newacronym{f2}{F2}{\textit{F2 score}} \newacronym{fbeta}{F-beta}{\textit{F-beta score}} \newacronym{mcc}{MCC}{Matthews \textit{correlation coefficient}} \newacronym{iou}{IoU}{\textit{intersection over union}} \newacronym{jaccard}{Jaccard}{Jaccard \textit{index}} \newacronym{hausdorff}{Hausdorff}{Hausdorff \textit{distance}} \newacronym{mse}{MSE}{\textit{mean squared error}} \newacronym{rmse}{RMSE}{\textit{root mean squared error}} \newacronym{mae}{MAE}{\textit{mean absolute error}} \newacronym{mape}{MAPE}{\textit{mean absolute percentage error}} \newacronym{smape}{SMAPE}{\textit{symmetric mean absolute percentage error}} \newacronym{r2}{R²}{\textit{coefficient of determination}} \newacronym{adjusted-r2}{Adjusted R²}{\textit{adjusted coefficient of determination}} \newacronym{aic}{AIC}{Akaike \textit{information criterion}} \newacronym{bic}{BIC}{Bayesian \textit{information criterion}} \newacronym{cv}{CV}{\textit{cross-validation}} \newacronym{loocv}{LOOCV}{\textit{leave-one-out cross-validation}} \newacronym{kfold}{k-fold}{\textit{k-fold cross-validation}} \newacronym{smote}{SMOTE}{\textit{Synthetic Minority Over-sampling Technique}} \newacronym{adasyn}{ADASYN}{\textit{Adaptive Synthetic Sampling}} \newacronym{enn}{ENN}{\textit{edited nearest neighbors}} \newacronym{ncr}{NCR}{\textit{neighborhood cleaning rule}} \newacronym{rus}{RUS}{\textit{random under-sampling}} \newacronym{ros}{ROS}{\textit{random over-sampling}} \newacronym{bce}{BCE}{\textit{binary cross-entropy}} \newacronym{cce}{CCE}{\textit{categorical cross-entropy}} \newacronym{scce}{SCCE}{\textit{sparse categorical cross-entropy}} \newacronym{nll}{NLL}{\textit{negative log-likelihood}} \newacronym{huber}{Huber}{Huber \textit{loss}} \newacronym{focal}{Focal}{\textit{focal loss}} \newacronym{dice-loss}{Dice Loss}{Dice \textit{loss}} \newacronym{tversky}{Tversky}{Tversky \textit{loss}} \newacronym{lovasz}{Lovász}{Lovász \textit{loss}} \newacronym{bn}{BN}{\textit{batch normalization}} \newacronym{ln}{LN}{\textit{layer normalization}} \newacronym{in}{IN}{\textit{instance normalization}} \newacronym{gn}{GN}{\textit{group normalization}} \newacronym{wn}{WN}{\textit{weight normalization}} \newacronym{sn}{SN}{\textit{spectral normalization}} \newacronym{rmsnorm}{RMSNorm}{\textit{root mean square layer normalization}} \newacronym{adain}{AdaIN}{\textit{adaptive instance normalization}} \newacronym{adagn}{AdaGN}{\textit{adaptive group normalization}} \newacronym{spade}{SPADE}{\textit{Spatially-Adaptive Normalization}} \newacronym{radam}{RAdam}{\textit{Rectified Adam}} \newacronym{lamb}{LAMB}{\textit{Layer-wise Adaptive Moments optimizer for Batch training}} \newacronym{lars}{LARS}{\textit{Layer-wise Adaptive Rate Scaling}} \newacronym{lookahead}{Lookahead}{Lookahead \textit{optimizer}} \newacronym{swats}{SWATS}{\textit{Switching from Adam to SGD}} \newacronym{slr}{SLR}{\textit{stochastic learning rate}} \newacronym{tta}{TTA}{\textit{test-time augmentation}} \newacronym{ema}{EMA}{\textit{exponential moving average}} \newacronym{swa}{SWA}{\textit{stochastic weight averaging}} \newacronym{ptq}{PTQ}{\textit{post-training quantization}} \newacronym{qat}{QAT}{\textit{quantization-aware training}} \newacronym{int8}{INT8}{\textit{8-bit integer}} \newacronym{fp16}{FP16}{\textit{16-bit floating point}} \newacronym{bf16}{BF16}{\textit{bfloat16}} \newacronym{fp32}{FP32}{\textit{32-bit floating point}} \newacronym{amp}{AMP}{\textit{automatic mixed precision}} \newacronym{ode}{ODE}{\textit{ordinary differential equation}} \newacronym{pde}{PDE}{\textit{partial differential equation}} \newacronym{sde}{SDE}{\textit{stochastic differential equation}} \newacronym{node}{NODE}{\textit{neural ordinary differential equation}} \newacronym{anode}{ANODE}{\textit{augmented neural ordinary differential equation}} \newacronym{pinn}{PINN}{\textit{physics-informed neural network}} \newacronym{deeponet}{DeepONet}{\textit{Deep Operator Network}} \newacronym{gnn}{GNN}{\textit{graph neural network}} \newacronym{gcn}{GCN}{\textit{graph convolutional network}} \newacronym{gat}{GAT}{\textit{graph attention network}} \newacronym{gin}{GIN}{\textit{Graph Isomorphism Network}} \newacronym{mpnn}{MPNN}{\textit{message passing neural network}} \newacronym{gatedgcn}{Gated GCN}{\textit{gated graph convolutional network}} \newacronym{line}{LINE}{\textit{Large-scale Information Network Embedding}} \newacronym{gran}{GRAN}{\textit{Graph Recurrent Attention Network}} \newacronym{dgmg}{DGMG}{\textit{Deep Generative Model of Graphs}} \newacronym{molgan}{MolGAN}{\textit{Molecular Generative Adversarial Network}} \newacronym{jtvae}{JT-VAE}{\textit{Junction Tree Variational Autoencoder}} \newacronym{rgcn}{R-GCN}{\textit{Relational Graph Convolutional Network}} \newacronym{hgt}{HGT}{\textit{Heterogeneous Graph Transformer}} \newacronym{han}{HAN}{\textit{Heterogeneous Graph Attention Network}} \newacronym{rgat}{R-GAT}{\textit{Relational Graph Attention Network}} \newacronym{compgcn}{CompGCN}{\textit{Composition-based Graph Convolutional Network}} \newacronym{kg}{KG}{\textit{knowledge graph}} \newacronym{kge}{KGE}{\textit{knowledge graph embedding}} \newacronym{kgc}{KGC}{\textit{knowledge graph completion}} \newacronym{qa}{QA}{\textit{question answering}} \newacronym{vqa}{VQA}{\textit{visual question answering}} \newacronym{squad}{SQuAD}{\textit{Stanford Question Answering Dataset}} \newacronym{glue}{GLUE}{\textit{General Language Understanding Evaluation}} \newacronym{superglue}{SuperGLUE}{\textit{Super General Language Understanding Evaluation}} \newacronym{xnli}{XNLI}{\textit{Cross-lingual Natural Language Inference}} \newacronym{mnli}{MNLI}{\textit{Multi-Genre Natural Language Inference}} \newacronym{snli}{SNLI}{Stanford \textit{Natural Language Inference}} \newacronym{qqp}{QQP}{Quora Question Pairs} \newacronym{qnli}{QNLI}{\textit{Question Natural Language Inference}} \newacronym{sts}{STS}{\textit{Semantic Textual Similarity}} \newacronym{cola}{CoLA}{\textit{Corpus of Linguistic Acceptability}} \newacronym{sst}{SST}{Stanford \textit{Sentiment Treebank}} \newacronym{mrpc}{MRPC}{Microsoft Research Paraphrase Corpus} \newacronym{rte}{RTE}{\textit{Recognizing Textual Entailment}} \newacronym{wnli}{WNLI}{Winograd \textit{Natural Language Inference}} \newacronym{wsc}{WSC}{Winograd \textit{Schema Challenge}} \newacronym{copa}{COPA}{\textit{Choice of Plausible Alternatives}} \newacronym{piqa}{PIQA}{\textit{Physical Interaction Question Answering}} \newacronym{siqa}{SIQA}{\textit{Social Interaction Question Answering}} \newacronym{commonsenseqa}{CommonsenseQA}{CommonsenseQA} \newacronym{arc}{ARC}{\textit{AI2 Reasoning Challenge}} \newacronym{race}{RACE}{\textit{ReAding Comprehension from Examinations}} \newacronym{drop}{DROP}{\textit{Discrete Reasoning Over Paragraphs}} \newacronym{hotpotqa}{HotpotQA}{HotpotQA} \newacronym{naturalquestions}{Natural Questions}{Natural Questions} \newacronym{triviaqa}{TriviaQA}{TriviaQA} \newacronym{quac}{QuAC}{\textit{Question Answering in Context}} \newacronym{coqa}{CoQA}{\textit{Conversational Question Answering}} \newacronym{multirc}{MultiRC}{\textit{Multi-Sentence Reading Comprehension}} \newacronym{record}{ReCoRD}{\textit{Reading Comprehension with Commonsense Reasoning Dataset}} \newacronym{wmt}{WMT}{\textit{Workshop on Machine Translation}} \newacronym{iwslt}{IWSLT}{\textit{International Workshop on Spoken Language Translation}}