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

Recently AddedMost HypedMost ActiveNeeds VerificationMost Verified

Showing 51–100 of 1978 papers

TitleStatusHype
MathOdyssey: Benchmarking Mathematical Problem-Solving Skills in Large Language Models Using Odyssey Math DataCode2
Unified-IO 2: Scaling Autoregressive Multimodal Models with Vision, Language, Audio, and ActionCode2
MASSIVE: A 1M-Example Multilingual Natural Language Understanding Dataset with 51 Typologically-Diverse LanguagesCode2
MetaMath: Bootstrap Your Own Mathematical Questions for Large Language ModelsCode2
Reflection-Tuning: Data Recycling Improves LLM Instruction-TuningCode2
SciAssess: Benchmarking LLM Proficiency in Scientific Literature AnalysisCode2
CLUE: A Chinese Language Understanding Evaluation BenchmarkCode2
The First Place Solution of WSDM Cup 2024: Leveraging Large Language Models for Conversational Multi-Doc QACode2
LoRA-Pro: Are Low-Rank Adapters Properly Optimized?Code2
Learning Transferable Visual Models From Natural Language SupervisionCode2
LoRI: Reducing Cross-Task Interference in Multi-Task Low-Rank AdaptationCode2
JGLUE: Japanese General Language Understanding EvaluationCode2
An Empirical Study of Qwen3 QuantizationCode2
BHASA: A Holistic Southeast Asian Linguistic and Cultural Evaluation Suite for Large Language ModelsCode2
BEVDriver: Leveraging BEV Maps in LLMs for Robust Closed-Loop DrivingCode2
GPT Understands, TooCode2
Encoder vs Decoder: Comparative Analysis of Encoder and Decoder Language Models on Multilingual NLU TasksCode2
An empirical study of LLaMA3 quantization: from LLMs to MLLMsCode2
DialogStudio: Towards Richest and Most Diverse Unified Dataset Collection for Conversational AICode2
DeBERTa: Decoding-enhanced BERT with Disentangled AttentionCode2
DeBERTaV3: Improving DeBERTa using ELECTRA-Style Pre-Training with Gradient-Disentangled Embedding SharingCode2
I-BERT: Integer-only BERT QuantizationCode2
Make LoRA Great Again: Boosting LoRA with Adaptive Singular Values and Mixture-of-Experts Optimization AlignmentCode2
It's Not Just Size That Matters: Small Language Models Are Also Few-Shot LearnersCode2
Large Language Model Safety: A Holistic SurveyCode2
Large Language Models Meet Knowledge Graphs for Question Answering: Synthesis and OpportunitiesCode2
Adversarial Regularization as Stackelberg Game: An Unrolled Optimization ApproachCode1
Visually-Situated Natural Language Understanding with Contrastive Reading Model and Frozen Large Language ModelsCode1
Cross-Lingual Semantic Role Labeling with High-Quality Translated Training CorpusCode1
Adversarial GLUE: A Multi-Task Benchmark for Robustness Evaluation of Language ModelsCode1
AutoTrans: Automating Transformer Design via Reinforced Architecture SearchCode1
CSKG: The CommonSense Knowledge GraphCode1
Automatic Disfluency Detection from Untranscribed SpeechCode1
Advances of Transformer-Based Models for News Headline GenerationCode1
CPT: A Pre-Trained Unbalanced Transformer for Both Chinese Language Understanding and GenerationCode1
Attention Is (not) All You Need for Commonsense ReasoningCode1
A Win-win Deal: Towards Sparse and Robust Pre-trained Language ModelsCode1
Convolution-enhanced Evolving Attention NetworksCode1
CREAK: A Dataset for Commonsense Reasoning over Entity KnowledgeCode1
C-STS: Conditional Semantic Textual SimilarityCode1
A Top-Down Neural Architecture towards Text-Level Parsing of Discourse Rhetorical StructureCode1
ATCO2 corpus: A Large-Scale Dataset for Research on Automatic Speech Recognition and Natural Language Understanding of Air Traffic Control CommunicationsCode1
ConFiguRe: Exploring Discourse-level Chinese Figures of SpeechCode1
Compresso: Structured Pruning with Collaborative Prompting Learns Compact Large Language ModelsCode1
CONDAQA: A Contrastive Reading Comprehension Dataset for Reasoning about NegationCode1
Conic10K: A Challenging Math Problem Understanding and Reasoning DatasetCode1
A Surprisingly Robust Trick for Winograd Schema ChallengeCode1
AceGPT, Localizing Large Language Models in ArabicCode1
A Survey of RWKVCode1
COMET-ATOMIC 2020: On Symbolic and Neural Commonsense Knowledge GraphsCode1
Show:102550
← PrevPage 2 of 40Next →
All datasetsPDP60STREUSLELexGLUEDialoGLUE fewshotDialoGLUE fullGLUE

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1HNNAccuracy90—Unverified
2UDSSM-II (ensemble)Accuracy78.3—Unverified
3BERT-large 340MAccuracy78.3—Unverified
4UDSSM-I (ensemble)Accuracy76.7—Unverified
5DSSMAccuracy75—Unverified
6UDSSM-IIAccuracy75—Unverified
7BERT-base 110M + MASAccuracy68.3—Unverified
8USSM + Supervised Deepnet + 3 Knowledge BasesAccuracy66.7—Unverified
9Word-level CNN+LSTM (full scoring)Accuracy60—Unverified
10Subword-level Transformer LMAccuracy58.3—Unverified
#ModelMetricClaimedVerifiedStatus
1BERT (pred POS/lemmas)Tags (Full) Acc82.5—Unverified
2BERT (none)Tags (Full) Acc82—Unverified
3BERT (gold POS/lemmas)Tags (Full) Acc81—Unverified
4GloVe (gold POS/lemmas)Tags (Full) Acc79.3—Unverified
5RoBERTa + LinearFull F1 (Preps)78.2—Unverified
6GloVe (none)Tags (Full) Acc77.5—Unverified
7GloVe (pred POS/lemmas)Tags (Full) Acc77.1—Unverified
8SVM (feature-rich, gold syntax)Role F1 (Preps)62.2—Unverified
9BiLSTM + MLP (gold syntax)Role F1 (Preps)62.2—Unverified
10SVM (feature-rich, auto syntax)Role F1 (Preps)58.2—Unverified
#ModelMetricClaimedVerifiedStatus
1CaseLaw-BERTCaseHOLD75.6—Unverified
2Legal-BERTCaseHOLD75.1—Unverified
3DeBERTaCaseHOLD72.1—Unverified
4LongformerCaseHOLD72—Unverified
5RoBERTaCaseHOLD71.7—Unverified
6BERTCaseHOLD70.7—Unverified
7BigBirdCaseHOLD70.4—Unverified
#ModelMetricClaimedVerifiedStatus
1ConvBERT-DGAverage74.6—Unverified
2ConvBERT-DG + Pre + MultiAverage73.8—Unverified
3mslmAverage73.49—Unverified
4ConvBERT + Pre + MultiAverage68.22—Unverified
5BanLanGenAverage39.16—Unverified
#ModelMetricClaimedVerifiedStatus
1ConvBERT + Pre + MultiAverage86.89—Unverified
2mslmAverage85.83—Unverified
3ConvBERT-DG + Pre + MultiAverage85.34—Unverified
#ModelMetricClaimedVerifiedStatus
1MT-DNN-SMARTAverage89.9—Unverified
2BERT-LARGEAverage82.1—Unverified