The Language Interpretability Tool: Extensible, Interactive Visualizations and Analysis for NLP Models

Tenney, Ian; Wexler, James; Bastings, Jasmijn; Bolukbasi, Tolga; Coenen, Andy; Gehrmann, Sebastian; Jiang, Ellen; Pushkarna, Mahima; Radebaugh, Carey; Reif, Emily; Yuan, Ann

doi:10.18653/v1/2020.emnlp-demos.15

Cited by 107 publications

(67 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Enriching ExplainaBoard with Glass-box Analysis EXPLAINABOARD currently performs blackbox analysis, solely analyzing system outputs without accessing model internals. On the other hand, there are many other glass-box interpretability tools that look at model internals, such as the AllenNLP Interpret (Wallace et al, 2019) and Language Interpretability Tool (Tenney et al, 2020). Expanding leaderboards to glass-box analysis methods (see Lipton (2018); Belinkov and Glass (2019) for a survey) is an interesting future work.…”

Section: Implications and Roadmapmentioning

confidence: 99%

ExplainaBoard: An Explainable Leaderboard for NLP

Liu

Xiao³

et al. 2021

Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Confer

View full text Add to dashboard Cite

With the rapid development of NLP research, leaderboards have emerged as one tool to track the performance of various systems on various NLP tasks. They are effective in this goal to some extent, but generally present a rather simplistic one-dimensional view of the submitted systems, communicated only through holistic accuracy numbers. In this paper, we present a new conceptualization and implementation of NLP evaluation: the EXPLAIN-ABOARD, which in addition to inheriting the functionality of the standard leaderboard, also allows researchers to (i) diagnose strengths and weaknesses of a single system (e.g. what is the best-performing system bad at?) (ii) interpret relationships between multiple systems. (e.g. where does system A outperform system B? What if we combine systems A, B and C?) and (iii) examine prediction results closely (e.g. what are common errors made by multiple systems or in what contexts do particular errors occur?). So far, EXPLAINABOARD covers more than 400 systems, 50 datasets, 40 languages, and 12 tasks. 1 We not only released an online platform at the website 2 but also make our evaluation tool an API with MIT Licence at Github 3 and PyPi 4 that allows users to conveniently assess their models offline. We additionally release all output files from systems that we have run or collected to motivate "output-driven" research in the future.

show abstract

Section: Implications and Roadmapmentioning

confidence: 99%

ExplainaBoard: An Explainable Leaderboard for NLP

Liu

Xiao³

et al. 2021

Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Confer

View full text Add to dashboard Cite

show abstract

“…The Newsroom dataset visualization tool (Grusky et al, 2018) highlights n-grams in the summary that overlap with the source article. The LIT tool (Tenney et al, 2020) highlights words or characters that differ between reference and generated texts. However neither tool aligns (Yousef and Janicke, 2021) the matched text.…”

Section: Related Workmentioning

confidence: 99%

SummVis: Interactive Visual Analysis of Models, Data, and Evaluation for Text Summarization

Vig¹,

Kryściński²,

Goel³

et al. 2021

Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Confer

View full text Add to dashboard Cite

Novel neural architectures, training strategies, and the availability of large-scale corpora haven been the driving force behind recent progress in abstractive text summarization. However, due to the black-box nature of neural models, uninformative evaluation metrics, and scarce tooling for model and data analysis, the true performance and failure modes of summarization models remain largely unknown. To address this limitation, we introduce SUMMVIS, an open-source tool for visualizing abstractive summaries that enables fine-grained analysis of the models, data, and evaluation metrics associated with text summarization. Through its lexical and semantic visualizations, the tools offers an easy entry point for in-depth model prediction exploration across important dimensions such as factual consistency or abstractiveness. The tool together with several pre-computed model outputs is available at https://summvis.com.

show abstract

“…Existing visualization systems and techniques do not visually connect attention mechanisms to linguistic knowledge (Tenney et al, 2020;DeRose et al, 2021), we propose novel visualization approaches that foster exploration across semantically and syntactically significant attention heads in complex model architectures. For example, for every attention head in the 144 heads of BERT, the entry A i,j in the attention map A, represents the attention weight from token i to token j.…”

Section: Introductionmentioning

confidence: 99%

Dodrio: Exploring Transformer Models with Interactive Visualization

Wang¹,

Turko²,

Chau³

2021

Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Confer

View full text Add to dashboard Cite

Figure 1: The DODRIO user interface showing user exploration of connections between attention weights from a fine-tuned BERT model and syntactic dependencies as well as semantic saliency scores on the SST2 dataset. (A) Dependency View enables users to hover over a word from the input sentence to highlight its associated dependency directed links as orange arcs (lighter is source; darker is target). (B) Semantic Attention Graph highlights the word's related tokens and their attentions; nodes are tokens (darker means more salient); a directed edge encodes attention weight between two tokens. (C) Attention Head Overview shows all attention heads in a multi-layer and multi-head model as a grid of circles, each head is (D) colored based on its linguistic knowledge in the model (more red→more semantic-aligned, more blue→more syntactic-aligned; darker→more aligned), and sized based on its importance score in the model (larger→more important).

show abstract

The Language Interpretability Tool: Extensible, Interactive Visualizations and Analysis for NLP Models

Cited by 107 publications

References 29 publications

ExplainaBoard: An Explainable Leaderboard for NLP

ExplainaBoard: An Explainable Leaderboard for NLP

SummVis: Interactive Visual Analysis of Models, Data, and Evaluation for Text Summarization

Dodrio: Exploring Transformer Models with Interactive Visualization

Contact Info

Product

Resources

About