Thailand

Larger language models have higher accuracy on average, but are they better on every single instance (datapoint)? Some work suggests larger models have higher out-of-distribution robustness, while other work suggests they have lower accuracy on rare subgroups. To understand these differences, we investigate these models at the level of individual instances. However, one major challenge is that individual predictions are highly sensitive to noise in the randomness in training. We develop statistically rigorous methods to address this, and after accounting for pretraining and finetuning noise, we find that our BERT-Large is worse than BERT-Mini on at least 1-4% of instances across MNLI, SST-2, and QQP, compared to the overall accuracy improvement of 2-10%. We also find that finetuning noise increases with model size, and that instance-level accuracy has momentum: improvement from BERT-Mini to BERT-Medium correlates with improvement from BERT-Medium to BERT-Large. Our findings suggest that instance-level predictions provide a rich source of information; we therefore recommend that researchers supplement model weights with model predictions.

IJCNLP-AACL 2021

Are Larger Pretrained Language Models Uniformly Better? Comparing Performance at the Instance Level

pre-training; analysis; statistics

**Welcome to ACL-IJCNLP 2021!**

The great event is jointly organized by the Association for Computational Linguistics (ACL) and Asian Federation of Natural Language Processing (AFNLP). 

As in previous years, the program of the conference includes a poster session, tutorials, workshops and demonstrations in addition to the main conference.


We were able to keep the registration fees similar to those charged for the virtual ACL 2020. The one fee allows attendance at the main conference and any/all tutorials and workshops. These fees would be $125 Regular Early and $175 Regular Late; $50 Student Early and $75 Student Late. Early registration closes at midnight July 11, 2021 (Eastern Daylight Time).

**Reminder:** It is ACL’s policy that at least one author of each accepted paper (including ACL Finding papers) must register for the conference.

**Reminder2:** Underline site will open closer to the event. If you already registered you will receive access detail

Registration is now open

The great event is jointly organized by the Association for Computational Linguistics (ACL) and Asian Federation of Natural Language Processing (AFNLP).

technical paper

Context-aware Adversarial Training for Name Regularity Bias in Named Entity Recognition

The scarcity of comprehensive up-to-date studies on evaluation metrics for text summarization and the lack of consensus regarding evaluation protocols continue to inhibit progress. We address the existing shortcomings of summarization evaluation methods along five dimensions: 1) we re-evaluate 14 automatic evaluation metrics in a comprehensive and consistent fashion using neural summarization model outputs along with expert and crowd-sourced human annotations; 2) we consistently benchmark 23 recent summarization models using the aforementioned automatic evaluation metrics; 3) we assemble the largest collection of summaries generated by models trained on the CNN/DailyMail news dataset and share it in a unified format; 4) we implement and share a toolkit that provides an extensible and unified API for evaluating summarization models across a broad range of automatic metrics; and 5) we assemble and share the largest and most diverse, in terms of model types, collection of human judgments of model-generated summaries on the CNN/Daily Mail dataset annotated by both expert judges and crowd-source workers. We hope that this work will help promote a more complete evaluation protocol for text summarization as well as advance research in developing evaluation metrics that better correlate with human judgments.

SummEval: Re-evaluating Summarization Evaluation

A desirable property of a reference-based evaluation metric that measures the content quality of a summary is that it should estimate how much information that summary has in common with a reference. Traditional text overlap based metrics such as ROUGE fail to achieve this because they are limited to matching tokens, either lexically or via embeddings. In this work, we propose a metric to evaluate the content quality of a summary using question-answering (QA). QA-based methods directly measure a summary's information overlap with a reference, making them fundamentally different than text overlap metrics. We demonstrate the experimental benefits of QA-based metrics through an analysis of our proposed metric, QAEval. QAEval out-performs current state-of-the-art metrics on most evaluations using benchmark datasets, while being competitive on others due to limitations of state-of-the-art models. Through a careful analysis of each component of QAEval, we identify its performance bottlenecks and estimate that its potential upper-bound performance surpasses all other automatic metrics, approaching that of the gold-standard Pyramid Method.

Towards Question-Answering as an Automatic Metric for Evaluating the Content Quality of a Summary

We propose Generation-Augmented Retrieval (GAR) for answering open-domain questions, which augments a query through text generation of heuristically discovered relevant contexts without external resources as supervision. 
We demonstrate that the generated contexts substantially enrich the semantics of the queries and GAR with sparse representations (BM25) achieves comparable or better performance than state-of-the-art dense retrieval methods such as DPR.
We show that generating diverse contexts for a query is beneficial as fusing their results consistently yields better retrieval accuracy.
Moreover, as sparse and dense representations are often complementary, GAR can be easily combined with DPR to achieve even better performance.
GAR achieves state-of-the-art performance on Natural Questions and TriviaQA datasets under the extractive QA setup when equipped with an extractive reader, and consistently outperforms other retrieval methods when the same generative reader is used.

Generation-Augmented Retrieval for Open-Domain Question Answering

Weakly supervised question answering usually has only the final answers as supervision signals while the correct solutions to derive the answers are not provided. This setting gives rise to the spurious solution problem: there may exist many spurious solutions that coincidentally derive the correct answer, but training on such solutions can hurt model performance (e.g., producing wrong solutions or answers). For example, for discrete reasoning tasks as on DROP, there may exist many equations to derive a numeric answer, and typically only one of them is correct. Previous learning methods mostly filter out spurious solutions with heuristics or using model confidence, but do not explicitly exploit the semantic correlations between a question and its solution. In this paper, to alleviate the spurious solution problem, we propose to explicitly exploit such semantic correlations by maximizing the mutual information between question-answer pairs and predicted solutions. Extensive experiments on four question answering datasets show that our method significantly outperforms previous learning methods in terms of task performance and is more effective in training models to produce correct solutions.

A Mutual Information Maximization Approach for the Spurious Solution Problem in Weakly Supervised Question Answering

Question answering (QA) in English has been widely explored, but multilingual datasets are relatively new, with several methods attempting to bridge the gap between high- and low-resourced languages using data augmentation through translation and cross-lingual transfer. In this project we take a step back and study which approaches allow us to take the most advantage of existing resources in order to produce QA systems in many languages. Specifically, we perform extensive analysis to measure the efficacy of few-shot approaches augmented with automatic translations and permutations of context-question-answer pairs. In addition, we make suggestions for future dataset development efforts that make better use of a fixed annotation budget, with a goal of increasing the language coverage of QA datasets and systems.

Towards more equitable question answering systems: How much more data do you need?

With the ever-increasing complexity of neural language models, practitioners have turned to methods for understanding the predictions of these models. One of the most well-adopted approaches for model interpretability is feature-based interpretability, i.e., ranking the features in terms of their impact on model predictions. Several prior studies have focused on assessing the fidelity of feature-based interpretability methods, i.e., measuring the impact of dropping the top-ranked features on the model output. However, relatively little work has been conducted on quantifying the robustness of interpretations. In this work, we assess the robustness of interpretations of neural text classifiers, specifically, those based on pretrained Transformer encoders, using two randomization tests. The first compares the interpretations of two models that are identical except for their initializations. The second measures whether the interpretations differ between a model with trained parameters and a model with random parameters. Both tests show surprising deviations from expected behavior, raising questions about the extent of insights that practitioners may draw from interpretations.

On the Lack of Robust Interpretability of Neural Text Classifiers

We introduce a new approach for the task of Controllable Text Simplification, where systems rewrite a complex English sentence so that it can be understood by readers at different grade levels in the US K-12 system. It uses a non-autoregressive model to iteratively edit an input sequence and incorporates lexical complexity information seamlessly into the refinement process to generate simplifications that better match the desired output complexity than strong autoregressive baselines. The analysis shows that our model's local edit operations are combined to achieve more complex simplification operations such as content deletion and paraphrasing, as well as sentence splitting.

A Non-Autoregressive Edit-Based Approach to Controllable Text Simplification

Context guides comprehenders’ expectations during language processing, and information-theoretic surprisal is commonly used as an index of cognitive processing effort. However, prior work using surprisal has considered only within-sentence context, using n-grams, neural language models, or syntactic structure as conditioning context. In this paper, we extend the surprisal approach to use broader topical context, investigating the influence of local and topical context on processing via an analysis of fMRI time courses collected during naturalistic listening. Lexical surprisal calculated from an LSTM language model is used to capture effects of local context; to capture the effects of broader context a new metric based on topic models, topical surprisal, is introduced. We identify distinct patterns of neural activation for lexical surprisal and topical surprisal. These differing neuro-anatomical correlates suggest that local and broad contextual cues during sentence processing recruit different brain regions and suggest that those regions of the language network functionally contribute to processing different dimensions of contextual information during comprehension. More generally, our approach adds to a growing literature using methods from computational linguistics to operationalize and test hypotheses about neuro-cognitive mechanisms in sentence processing.

Using surprisal and fMRI to map the neural bases of broad and local contextual prediction during natural language comprehension

Multilingual Transformer-based language models, usually pretrained on more than 100 languages, have been shown to achieve outstanding results in a wide range of cross-lingual transfer tasks. However, it remains unknown whether the optimization for different languages conditions the capacity of the models to generalize over syntactic structures, and how languages with syntactic phenomena of different complexity are affected. In this work, we explore the syntactic generalization capabilities of the monolingual and multilingual versions of BERT and RoBERTa. 
More specifically, we evaluate the syntactic generalization potential of the models on English and Spanish tests, comparing the syntactic abilities of monolingual and multilingual models on the same language (English), and of multilingual models on two different languages (English and Spanish). For English, we use the available SyntaxGym test suite; for Spanish, we introduce SyntaxGymES, a novel ensemble of targeted syntactic tests in Spanish, designed to evaluate the syntactic generalization capabilities of language models through the SyntaxGym online platform.

Are Larger Pretrained Language Models Uniformly Better? Comparing Performance at the Instance Level

Downloads

Next from IJCNLP-AACL 2021

Context-aware Adversarial Training for Name Regularity Bias in Named Entity Recognition

Similar lecture

Guiding the Growth: Difficulty-Controllable Question Generation through Step-by-Step Rewriting

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES

Are Larger Pretrained Language Models Uniformly Better? Comparing Performance at the Instance Level

.css-70qvj9{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;}Downloads

Next from IJCNLP-AACL 2021

Context-aware Adversarial Training for Name Regularity Bias in Named Entity Recognition

Similar lecture

Guiding the Growth: Difficulty-Controllable Question Generation through Step-by-Step Rewriting

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES

Downloads