SOTAVerified

Natural Language Understanding

Natural Language Understanding is an important field of Natural Language Processing which contains various tasks such as text classification, natural language inference and story comprehension. Applications enabled by natural language understanding range from question answering to automated reasoning.

Source: Find a Reasonable Ending for Stories: Does Logic Relation Help the Story Cloze Test?

Papers

Showing 101150 of 1978 papers

TitleStatusHype
Joint Prompt Optimization of Stacked LLMs using Variational InferenceCode1
Debiasing Methods in Natural Language Understanding Make Bias More AccessibleCode1
AceGPT, Localizing Large Language Models in ArabicCode1
Distilling Large Vision-Language Model with Out-of-Distribution GeneralizabilityCode1
DroidCall: A Dataset for LLM-powered Android Intent InvocationCode1
Analyzing Multi-Task Learning for Abstractive Text SummarizationCode1
Dual Supervised Learning for Natural Language Understanding and GenerationCode1
An Annotated Corpus of Reference Resolution for Interpreting Common GroundingCode1
DialoGLUE: A Natural Language Understanding Benchmark for Task-Oriented DialogueCode1
A Dataset for Statutory Reasoning in Tax Law Entailment and Question AnsweringCode1
A Surprisingly Robust Trick for Winograd Schema ChallengeCode1
ELECTRA: Pre-training Text Encoders as Discriminators Rather Than GeneratorsCode1
A Survey of RWKVCode1
Empowering News Recommendation with Pre-trained Language ModelsCode1
Does It Make Sense? And Why? A Pilot Study for Sense Making and ExplanationCode1
An Empirical Study of Tokenization Strategies for Various Korean NLP TasksCode1
End-to-End Slot Alignment and Recognition for Cross-Lingual NLUCode1
Enhancing Language Representation with Constructional Information for Natural Language UnderstandingCode1
Adversarial GLUE: A Multi-Task Benchmark for Robustness Evaluation of Language ModelsCode1
data2vec: A General Framework for Self-supervised Learning in Speech, Vision and LanguageCode1
Evaluating Scoped Meaning RepresentationsCode1
Evaluating the Logical Reasoning Ability of ChatGPT and GPT-4Code1
Evolving Attention with Residual ConvolutionsCode1
FedEx-LoRA: Exact Aggregation for Federated and Efficient Fine-Tuning of Foundation ModelsCode1
Adversarial Regularization as Stackelberg Game: An Unrolled Optimization ApproachCode1
Explaining Patterns in Data with Language Models via Interpretable AutopromptingCode1
Efficient Self-supervised Learning with Contextualized Target Representations for Vision, Speech and LanguageCode1
FaVIQ: FAct Verification from Information-seeking QuestionsCode1
AllenNLP: A Deep Semantic Natural Language Processing PlatformCode1
Visually-Situated Natural Language Understanding with Contrastive Reading Model and Frozen Large Language ModelsCode1
CPT: A Pre-Trained Unbalanced Transformer for Both Chinese Language Understanding and GenerationCode1
A Simple but Tough-to-Beat Data Augmentation Approach for Natural Language Understanding and GenerationCode1
CREAK: A Dataset for Commonsense Reasoning over Entity KnowledgeCode1
Cross-Lingual Semantic Role Labeling with High-Quality Translated Training CorpusCode1
Conic10K: A Challenging Math Problem Understanding and Reasoning DatasetCode1
Adapting Pretrained Transformer to Lattices for Spoken Language UnderstandingCode1
ConvBERT: Improving BERT with Span-based Dynamic ConvolutionCode1
CONDAQA: A Contrastive Reading Comprehension Dataset for Reasoning about NegationCode1
ConFiguRe: Exploring Discourse-level Chinese Figures of SpeechCode1
Convolution-enhanced Evolving Attention NetworksCode1
CSKG: The CommonSense Knowledge GraphCode1
ALBETO and DistilBETO: Lightweight Spanish Language ModelsCode1
Collaborative Multi-Agent Dialogue Model Training Via Reinforcement LearningCode1
COMET-ATOMIC 2020: On Symbolic and Neural Commonsense Knowledge GraphsCode1
A Large-scale Comprehensive Abusiveness Detection Dataset with Multifaceted Labels from RedditCode1
CLUTRR: A Diagnostic Benchmark for Inductive Reasoning from TextCode1
Comparison by Conversion: Reverse-Engineering UCCA from Syntax and Lexical SemanticsCode1
CLINE: Contrastive Learning with Semantic Negative Examples for Natural Language UnderstandingCode1
AdaMix: Mixture-of-Adaptations for Parameter-efficient Model TuningCode1
A Sequence-to-Sequence Approach to Dialogue State TrackingCode1
Show:102550
← PrevPage 3 of 40Next →

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1HNNAccuracy90Unverified
2UDSSM-II (ensemble)Accuracy78.3Unverified
3BERT-large 340MAccuracy78.3Unverified
4UDSSM-I (ensemble)Accuracy76.7Unverified
5DSSMAccuracy75Unverified
6UDSSM-IIAccuracy75Unverified
7BERT-base 110M + MASAccuracy68.3Unverified
8USSM + Supervised Deepnet + 3 Knowledge BasesAccuracy66.7Unverified
9Word-level CNN+LSTM (full scoring)Accuracy60Unverified
10Subword-level Transformer LMAccuracy58.3Unverified
#ModelMetricClaimedVerifiedStatus
1BERT (pred POS/lemmas)Tags (Full) Acc82.5Unverified
2BERT (none)Tags (Full) Acc82Unverified
3BERT (gold POS/lemmas)Tags (Full) Acc81Unverified
4GloVe (gold POS/lemmas)Tags (Full) Acc79.3Unverified
5RoBERTa + LinearFull F1 (Preps)78.2Unverified
6GloVe (none)Tags (Full) Acc77.5Unverified
7GloVe (pred POS/lemmas)Tags (Full) Acc77.1Unverified
8SVM (feature-rich, gold syntax)Role F1 (Preps)62.2Unverified
9BiLSTM + MLP (gold syntax)Role F1 (Preps)62.2Unverified
10SVM (feature-rich, auto syntax)Role F1 (Preps)58.2Unverified
#ModelMetricClaimedVerifiedStatus
1CaseLaw-BERTCaseHOLD75.6Unverified
2Legal-BERTCaseHOLD75.1Unverified
3DeBERTaCaseHOLD72.1Unverified
4LongformerCaseHOLD72Unverified
5RoBERTaCaseHOLD71.7Unverified
6BERTCaseHOLD70.7Unverified
7BigBirdCaseHOLD70.4Unverified
#ModelMetricClaimedVerifiedStatus
1ConvBERT-DGAverage74.6Unverified
2ConvBERT-DG + Pre + MultiAverage73.8Unverified
3mslmAverage73.49Unverified
4ConvBERT + Pre + MultiAverage68.22Unverified
5BanLanGenAverage39.16Unverified
#ModelMetricClaimedVerifiedStatus
1ConvBERT + Pre + MultiAverage86.89Unverified
2mslmAverage85.83Unverified
3ConvBERT-DG + Pre + MultiAverage85.34Unverified
#ModelMetricClaimedVerifiedStatus
1MT-DNN-SMARTAverage89.9Unverified
2BERT-LARGEAverage82.1Unverified