Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)

By K. Erk, Noah A. Smith. 2016

1,401 citations recorded by the source corpus.

Subfields: Natural Language Processing, Evaluation & Benchmarks

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Natural Language Processing

In Evaluation & Benchmarks

Citations in the corpus

Using DeepSpeed and Megatron to Train Megatron-Turing NLG 530B, A Large-Scale Generative Language Model (2022, background): “Table 2: LAMBADA zero-shot, one-shot and few-shot accuracy.”
PEGASUS: Pre-training with Extracted Gap-sentences for Abstractive Summarization (2019, neutral): “As shown in Figure 5, BPE and Unigram were comparable on news datasets while Unigram outperformed BPE 2Implemented in https://github.com/google/sentencepiece on non-news datasets, especially WikiHow.”
An Empirical Study of Metrics to Measure Representational Harms in Pre-Trained Language Models (2023, methodology): “We measure the perplexity of student models on language modeling benchmarks including WikiText-2, WikiText-103 (Merity et al., 2016), Lambada (Paperno et al., 2016), and the Penn Treebank (Marcus et al., 1993) (Appendix A.6, Table 15).”
Evidence Aggregation for Answer Re-Ranking in Open-Domain Question Answering (2017, background): “(2) KB-QA systems (Bast & Haussmann, 2015; Yih et al., 2015; Xu et al., 2016) sometimes use text evidence to enhance answer re-ranking, where the features are also extracted on the pair of question and a single-passage but ignored the union information among multiple passages.”
Pre-train, Prompt, and Predict: A Systematic Survey of Prompting Methods in Natural Language Processing (2021, neutral): “Specifically, the guidance signals could be style tokens [27, 123], length specifications [56], domain tags [14], or any variety of other pieces of information used to control of the generated text.”
CommonsenseQA 2.0: Exposing the Limits of AI through Gamification (2021, background): “Unfortunately, the vast majority of existing benchmarks are based on crowdsourcing, which are by and large static and non-interactive, leading to datasets that are plagued with shortcuts and artifacts [18, 19, 20, 4].”
Cross-lingual Retrieval for Iterative Self-Supervised Training (2020, background): “In the supervised machine learning literature, data augmentation [5, 21, 41, 50] has been applied to improve learning performance.”
What Language Model Architecture and Pretraining Objective Work Best for Zero-Shot Generalization? (2022, background): “On non-T0-Eval tasks, in contrast, causal decoder performance shoots up dramatically, although a lot of the difference is driven by LAMBADA [Paperno et al., 2016a], a language modeling task.”
MERLOT RESERVE: Neural Script Knowledge through Vision and Language and Sound (2022, neutral): “[99] Rico Sennrich, Barry Haddow, and Alexandra Birch.”
MERLOT: Multimodal Neural Script Knowledge Models (2021, methodology): “To split the videos into segments, we byte-pair-encode (BPE) each video transcript and align the tokens with YouTube’s word-level timestamps [86, 76].”
Scaling Autoregressive Models for Content-Rich Text-to-Image Generation (2022, neutral): “For text encoding, we build a sentence-piece model (Sennrich et al., 2016; Kudo & Richardson, 2018) of vocabulary size 16,000 on a sampled text corpus from the training data (Section 4.”
Visualisation and 'diagnostic classifiers' reveal how recurrent and recursive neural networks process hierarchical structure (2017, neutral): “…tasks such as machine translation (e.g. Sutskever et al., 2014; Bahdanau et al., 2015), sentiment analysis (e.g. Kalchbrenner et al., 2014; Socher et al., 2013; Le and Zuidema, 2015), semantic role labelling (e.g. Roth and Lapata, 2016) and recognizing textual entailment (e.g. Bowman et al., 2016).”
Unsupervised Neural Machine Translation (2017, methodology): “C L ] 3 0 O ct 2 (Sennrich et al., 2016a) into the training procedure to further improve results.”
Performance Impact Caused by Hidden Bias of Training Data for Recognizing Textual Entailment (2018, methodology): “…important, because the SNLI corpus is widely used as training data by many NN models for RTE (Bowman et al., 2015; Rocktäschel et al., 2015; Yin et al., 2016; Mou et al., 2016a; Wang and Jiang, 2016; Liu et al., 2016a; Liu et al., 2016b; Cheng et al., 2016; Parikh et al., 2016; Sha et al., 2016).”