arXiv subjects

Steffen Eger

Publications and source records attributed to Steffen Eger.

At least 91 records · Page 5Linked to original sources

arXiv2021

BERT-Defense: A Probabilistic Model Based on BERT to Combat Cognitively Inspired Orthographic Adversarial Attacks

Adversarial attacks expose important blind spots of deep learning systems. While word- and sentence-level attack scenarios mostly deal with finding semantic paraphrases of the input that fool NLP models, character-level attacks typically insert typos into the input stream. It is commonly thought that these are easier to defend via spelling correction modules. In this work, we show that both a standard spellchecker and the approach of Pruthi et al. (2019), which trains to defend against insertions, deletions and swaps, perform poorly on the character-level benchmark recently proposed in Eger and Benz (2020) which includes more challenging attacks such as visual and phonetic perturbations and missing word segmentations. In contrast, we show that an untrained iterative approach which combines context-independent character-level information with context-dependent information from BERT's masked language modeling can perform on par with human crowd-workers from Amazon Mechanical Turk (AMT) supervised via 3-shot learning.

cs.CL↗

arXiv2020

Probing Multilingual BERT for Genetic and Typological Signals

We probe the layers in multilingual BERT (mBERT) for phylogenetic and geographic language signals across 100 languages and compute language distances based on the mBERT representations. We 1) employ the language distances to infer and evaluate language trees, finding that they are close to the reference family tree in terms of quartet tree distance, 2) perform distance matrix regression analysis, finding that the language distances can be best explained by phylogenetic and worst by structural factors and 3) present a novel measure for measuring diachronic meaning stability (based on cross-lingual representation variability) which correlates significantly with published ranked lists based on linguistic approaches. Our results contribute to the nascent field of typological interpretability of cross-lingual text representations.

cs.CL↗

arXiv2020

Vec2Sent: Probing Sentence Embeddings with Natural Language Generation

We introspect black-box sentence embeddings by conditionally generating from them with the objective to retrieve the underlying discrete sentence. We perceive of this as a new unsupervised probing task and show that it correlates well with downstream task performance. We also illustrate how the language generated from different encoders differs. We apply our approach to generate sentence analogies from sentence embeddings.

cs.CL↗

arXiv2020

How to Probe Sentence Embeddings in Low-Resource Languages: On Structural Design Choices for Probing Task Evaluation

Sentence encoders map sentences to real valued vectors for use in downstream applications. To peek into these representations - e.g., to increase interpretability of their results - probing tasks have been designed which query them for linguistic knowledge. However, designing probing tasks for lesser-resourced languages is tricky, because these often lack large-scale annotated data or (high-quality) dependency parsers as a prerequisite of probing task design in English. To investigate how to probe sentence embeddings in such cases, we investigate sensitivity of probing task results to structural design choices, conducting the first such large scale study. We show that design choices like size of the annotated probing dataset and type of classifier used for evaluation do (sometimes substantially) influence probing outcomes. We then probe embeddings in a multilingual setup with design choices that lie in a 'stable region', as we identify for English, and find that results on English do not transfer to other languages. Fairer and more comprehensive sentence-level probing evaluation should thus be carried out on multiple languages in the future.

cs.CL↗

arXiv2020

From Hero to Zéroe: A Benchmark of Low-Level Adversarial Attacks

Adversarial attacks are label-preserving modifications to inputs of machine learning classifiers designed to fool machines but not humans. Natural Language Processing (NLP) has mostly focused on high-level attack scenarios such as paraphrasing input texts. We argue that these are less realistic in typical application scenarios such as in social media, and instead focus on low-level attacks on the character-level. Guided by human cognitive abilities and human robustness, we propose the first large-scale catalogue and benchmark of low-level adversarial attacks, which we dub Zéroe, encompassing nine different attack modes including visual and phonetic adversaries. We show that RoBERTa, NLP's current workhorse, fails on our attacks. Our dataset provides a benchmark for testing robustness of future more human-like NLP models.

cs.CL↗

arXiv2020

Text Processing Like Humans Do: Visually Attacking and Shielding NLP Systems

Visual modifications to text are often used to obfuscate offensive comments in social media (e.g., "!d10t") or as a writing style ("1337" in "leet speak"), among other scenarios. We consider this as a new type of adversarial attack in NLP, a setting to which humans are very robust, as our experiments with both simple and more difficult visual input perturbations demonstrate. We then investigate the impact of visual adversarial attacks on current NLP systems on character-, word-, and sentence-level tasks, showing that both neural and non-neural models are, in contrast to humans, extremely sensitive to such attacks, suffering performance decreases of up to 82\%. We then explore three shielding methods---visual character embeddings, adversarial training, and rule-based recovery---which substantially improve the robustness of the models. However, the shielding methods still fall behind performances achieved in non-attack scenarios, which demonstrates the difficulty of dealing with visual attacks.

cs.CL↗

arXiv2020

On the Limitations of Cross-lingual Encoders as Exposed by Reference-Free Machine Translation Evaluation

Evaluation of cross-lingual encoders is usually performed either via zero-shot cross-lingual transfer in supervised downstream tasks or via unsupervised cross-lingual textual similarity. In this paper, we concern ourselves with reference-free machine translation (MT) evaluation where we directly compare source texts to (sometimes low-quality) system translations, which represents a natural adversarial setup for multilingual encoders. Reference-free evaluation holds the promise of web-scale comparison of MT systems. We systematically investigate a range of metrics based on state-of-the-art cross-lingual semantic representations obtained with pretrained M-BERT and LASER. We find that they perform poorly as semantic encoders for reference-free MT evaluation and identify their two key limitations, namely, (a) a semantic mismatch between representations of mutual translations and, more prominently, (b) the inability to punish "translationese", i.e., low-quality literal translations. We propose two partial remedies: (1) post-hoc re-alignment of the vector spaces and (2) coupling of semantic-similarity based metrics with target-side language modeling. In segment-level MT evaluation, our best metric surpasses reference-based BLEU by 5.7 correlation points.

cs.CL↗

arXiv2020

SUPERT: Towards New Frontiers in Unsupervised Evaluation Metrics for Multi-Document Summarization

We study unsupervised multi-document summarization evaluation metrics, which require neither human-written reference summaries nor human annotations (e.g. preferences, ratings, etc.). We propose SUPERT, which rates the quality of a summary by measuring its semantic similarity with a pseudo reference summary, i.e. selected salient sentences from the source documents, using contextualized embeddings and soft token alignment techniques. Compared to the state-of-the-art unsupervised evaluation metrics, SUPERT correlates better with human ratings by 18-39%. Furthermore, we use SUPERT as rewards to guide a neural-based reinforcement learning summarizer, yielding favorable performance compared to the state-of-the-art unsupervised summarizers. All source code is available at https://github.com/yg211/acl20-ref-free-eval.

cs.CL↗

arXiv2019

MoverScore: Text Generation Evaluating with Contextualized Embeddings and Earth Mover Distance

A robust evaluation metric has a profound impact on the development of text generation systems. A desirable metric compares system output against references based on their semantics rather than surface forms. In this paper we investigate strategies to encode system and reference texts to devise a metric that shows a high correlation with human judgment of text quality. We validate our new metric, namely MoverScore, on a number of text generation tasks including summarization, machine translation, image captioning, and data-to-text generation, where the outputs are produced by a variety of neural and non-neural systems. Our findings suggest that metrics combining contextualized representations with a distance measure perform the best. Such metrics also demonstrate strong generalization capability across tasks. For ease-of-use we make our metrics available as web service.

cs.CL↗

arXiv2019

Semantic Change and Emerging Tropes In a Large Corpus of New High German Poetry

Due to its semantic succinctness and novelty of expression, poetry is a great test bed for semantic change analysis. However, so far there is a scarcity of large diachronic corpora. Here, we provide a large corpus of German poetry which consists of about 75k poems with more than 11 million tokens, with poems ranging from the 16th to early 20th century. We then track semantic change in this corpus by investigating the rise of tropes (`love is magic') over time and detecting change points of meaning, which we find to occur particularly within the German Romantic period. Additionally, through self-similarity, we reconstruct literary periods and find evidence that the law of linear semantic change also applies to poetry.

cs.CL↗

arXiv2019

DBPal: Weak Supervision for Learning a Natural Language Interface to Databases

This paper describes DBPal, a new system to translate natural language utterances into SQL statements using a neural machine translation model. While other recent approaches use neural machine translation to implement a Natural Language Interface to Databases (NLIDB), existing techniques rely on supervised learning with manually curated training data, which results in substantial overhead for supporting each new database schema. In order to avoid this issue, DBPal implements a novel training pipeline based on weak supervision that synthesizes all training data from a given database schema. In our evaluation, we show that DBPal can outperform existing rule-based NLIDBs while achieving comparable performance to other NLIDBs that leverage deep neural network models without relying on manually curated training data for every new database schema.

cs.DB↗

arXiv2019

Towards Scalable and Reliable Capsule Networks for Challenging NLP Applications

Obstacles hindering the development of capsule networks for challenging NLP applications include poor scalability to large output spaces and less reliable routing processes. In this paper, we introduce: 1) an agreement score to evaluate the performance of routing processes at instance level; 2) an adaptive optimizer to enhance the reliability of routing; 3) capsule compression and partial routing to improve the scalability of capsule networks. We validate our approach on two NLP tasks, namely: multi-label text classification and question answering. Experimental results show that our approach considerably improves over strong competitors on both tasks. In addition, we gain the best results in low-resource settings with few training instances.

cs.CL↗

arXiv2019

Pitfalls in the Evaluation of Sentence Embeddings

Deep learning models continuously break new records across different NLP tasks. At the same time, their success exposes weaknesses of model evaluation. Here, we compile several key pitfalls of evaluation of sentence embeddings, a currently very popular NLP paradigm. These pitfalls include the comparison of embeddings of different sizes, normalization of embeddings, and the low (and diverging) correlations between transfer and probing tasks. Our motivation is to challenge the current evaluation of sentence embeddings and to provide an easy-to-access reference for future research. Based on our insights, we also recommend better practices for better future evaluations of sentence embeddings.

cs.CL↗

arXiv2019

Does My Rebuttal Matter? Insights from a Major NLP Conference

Peer review is a core element of the scientific process, particularly in conference-centered fields such as ML and NLP. However, only few studies have evaluated its properties empirically. Aiming to fill this gap, we present a corpus that contains over 4k reviews and 1.2k author responses from ACL-2018. We quantitatively and qualitatively assess the corpus. This includes a pilot study on paper weaknesses given by reviewers and on quality of author responses. We then focus on the role of the rebuttal phase, and propose a novel task to predict after-rebuttal (i.e., final) scores from initial reviews and author responses. Although author responses do have a marginal (and statistically significant) influence on the final scores, especially for borderline papers, our results suggest that a reviewer's final score is largely determined by her initial score and the distance to the other reviewers' initial scores. In this context, we discuss the conformity bias inherent to peer reviewing, a bias that has largely been overlooked in previous research. We hope our analyses will help better assess the usefulness of the rebuttal phase in NLP conferences.

cs.CL↗

arXiv2019

Predicting Research Trends From Arxiv

We perform trend detection on two datasets of Arxiv papers, derived from its machine learning (cs.LG) and natural language processing (cs.CL) categories. Our approach is bottom-up: we first rank papers by their normalized citation counts, then group top-ranked papers into different categories based on the tasks that they pursue and the methods they use. We then analyze these resulting topics. We find that the dominating paradigm in cs.CL revolves around natural language generation problems and those in cs.LG revolve around reinforcement learning and adversarial principles. By extrapolation, we predict that these topics will remain lead problems/approaches in their fields in the short- and mid-term.

cs.CL↗

arXiv2019

Is it Time to Swish? Comparing Deep Learning Activation Functions Across NLP tasks

Activation functions play a crucial role in neural networks because they are the nonlinearities which have been attributed to the success story of deep learning. One of the currently most popular activation functions is ReLU, but several competitors have recently been proposed or 'discovered', including LReLU functions and swish. While most works compare newly proposed activation functions on few tasks (usually from image classification) and against few competitors (usually ReLU), we perform the first large-scale comparison of 21 activation functions across eight different NLP tasks. We find that a largely unknown activation function performs most stably across all tasks, the so-called penalized tanh function. We also show that it can successfully replace the sigmoid and tanh gates in LSTM cells, leading to a 2 percentage point (pp) improvement over the standard choices on a challenging NLP task.

cs.CL↗

arXiv2018

Concatenated Power Mean Word Embeddings as Universal Cross-Lingual Sentence Representations

Average word embeddings are a common baseline for more sophisticated sentence embedding techniques. However, they typically fall short of the performances of more complex models such as InferSent. Here, we generalize the concept of average word embeddings to power mean word embeddings. We show that the concatenation of different types of power mean word embeddings considerably closes the gap to state-of-the-art methods monolingually and substantially outperforms these more complex techniques cross-lingually. In addition, our proposed method outperforms different recently proposed baselines such as SIF and Sent2Vec by a solid margin, thus constituting a much harder-to-beat monolingual baseline. Our data and code are publicly available.

cs.CL↗

arXiv2018

The Combinatorics of Weighted Vector Compositions

A vector composition of a vector $\mathbf{\ell}$ is a matrix $\mathbf{A}$ whose rows sum to $\mathbf{\ell}$. We define a weighted vector composition as a vector composition in which the column values of $\mathbf{A}$ may appear in different colors. We study vector compositions from different viewpoints: (1) We show how they are related to sums of random vectors and (2) how they allow to derive formulas for partial derivatives of composite functions. (3) We study congruence properties of the number of weighted vector compositions, for fixed and arbitrary number of parts, many of which are analogous to those of ordinary binomial coefficients and related quantities. Via the Central Limit Theorem and their multivariate generating functions, (4) we also investigate the asymptotic behavior of several special cases of numbers of weighted vector compositions. Finally, (5) we conjecture an extension of a primality criterion due to Mann and Shanks in the context of weighted vector compositions.

math.CO↗