Influence Tuning: Demoting Spurious Correlations via Instance Attribution and Instance-Driven Updates

Among the most critical limitations of deep learning NLP models are their lack of interpretability, and their reliance on spurious correlations. Prior work proposed various approaches to interpreting the black-box models to unveil the spurious correlations, but the research was primarily used in human-computer interaction scenarios. It still remains underexplored whether or how such model interpretations can be used to automatically unlearn'' confounding features. In this work, we propose influence tuning---a procedure that leverages model interpretations to update the model parameters towards a plausible interpretation (rather than an interpretation that relies on spurious patterns in the data) in addition to learning to predict the task labels. We show that in a controlled setup, influence tuning can help deconfounding the model from spurious patterns in data, significantly outperforming baseline methods that use adversarial training.

A Call for Clarity in Contemporary Authorship Attribution Evaluation

Recent research has documented that results reported in frequently-cited authorship attribution papers are difficult to reproduce. Inaccessible code and data are often proposed as factors which block successful reproductions. Even when original materials are available, problems remain which prevent researchers from comparing the effectiveness of different methods. To solve the remaining problems---the lack of fixed test sets and the use of inappropriately homogeneous corpora---our paper contributes materials for five closed-set authorship identification experiments. The five experiments feature texts from 106 distinct authors. Experiments involve a range of contemporary non-fiction American English prose. These experiments provide the foundation for comparable and reproducible authorship attribution research involving contemporary writing.

Small-Scale Cross-Language Authorship Attribution on Social Media Comments

Cross-language authorship attribution is the challenging task of classifying documents by bilingual authors where the training documents are written in a different language than the evaluation documents. Traditional solutions rely on either translation to enable the use of single-language features, or language-independent feature extraction methods. More recently, transformer-based language models like BERT can also be pre-trained on multiple languages, making them intuitive candidates for cross-language classifiers which have not been used for this task yet. We perform extensive experiments to benchmark the performance of three different approaches to a smallscale cross-language authorship attribution experiment: (1) using language-independent features with traditional classification models, (2) using multilingual pre-trained language models, and (3) using machine translation to allow single-language classification. For the language-independent features, we utilize universal syntactic features like part-of-speech tags and dependency graphs, and multilingual BERT as a pre-trained language model. We use a small-scale social media comments dataset, closely reflecting practical scenarios. We show that applying machine translation drastically increases the performance of almost all approaches, and that the syntactic features in combination with the translation step achieve the best overall classification performance. In particular, we demonstrate that pre-trained language models are outperformed by traditional models in small scale authorship attribution problems for every language combination analyzed in this paper.

Aligning Faithful Interpretations with their Social Attribution

Abstract We find that the requirement of model interpretations to be faithful is vague and incomplete. With interpretation by textual highlights as a case study, we present several failure cases. Borrowing concepts from social science, we identify that the problem is a misalignment between the causal chain of decisions (causal attribution) and the attribution of human behavior to the interpretation (social attribution). We reformulate faithfulness as an accurate attribution of causality to the model, and introduce the concept of aligned faithfulness: faithful causal chains that are aligned with their expected social behavior. The two steps of causal attribution and social attribution together complete the process of explaining behavior. With this formalization, we characterize various failures of misaligned faithful highlight interpretations, and propose an alternative causal chain to remedy the issues. Finally, we implement highlight explanations of the proposed causal format using contrastive explanations.

Casual Attribution and Its Relation with The Scholastic Achievement Motivation on A Sample of Superior and Ordinary Students in Lattakia Governorate

This research aimed at knowing the relation between casual attribution, and scholastic Achievement Motivation on a sample of superior And ordinary students, and at knowing the differences between the two sorts of each sample the research depended on the descriptive and analytic method by using casual attribution scale which has been designed by the researcher the sample was (288) students of males and females from the eighth grade at schools of superior And ordinary in Lattakia governorate: The results were as follows: There is a statistical significant positive correlative relation between attribution of (effort making) and scholastic Achievement Motivation of ordinary students students while there is no statistical significant correlative relation between the other casual attributions (the ability, the luck (chance), the task difficulty, (the mixed attribution) and scholastic Achievement Motivation. There is a statistical significant positive correlative relation between each of (the ability attribution, the effort attribution, the task difficulty attribution, and between scholastic Achievement Motivation of superior students, while there is no statistical significant correlative relation between, the luck attribution, and the mixed attribution, and scholastic Achievement Motivation. There are statistical significant differences between the mean scores of ordinary students responses and superior students ones according to casual attribution scale except for the mixed attribution, and there are differences according to scholastic Achievement Motivation scale in favor of the superior students. There are no a statistical significant differences between the mean scores of male students responses and female students ones according to both scales the casual attribution and scholastic Achievement Motivation.