Papers

Interpreting Models

• Information-Theoretic Probing with Minimum Description Length
• Designing and Interpreting Probes with Control Tasks
• Analysis Methods in Neural Language Processing: A Survey

Model Evaluation

• Beyond Accuracy: Behavioral Testing of NLP Models with CheckList
• Evaluating NLP Models via Contrast Sets
• BERTScore: Evaluating Text Generation with BERT
• BLEURT: Learning Robust Metrics for Text Generation
• MoverScore: Text Generation Evaluating with Contextualized Embeddings and Earth Mover Distance

Robustness

• Learning the Difference that Makes a Difference with Counterfactually-Augmented Data
• Certified Robustness to Adversarial Word Substitutions
• Universal Adversarial Triggers for Attacking and Analyzing NLP
• Seq2Sick: Evaluating the Robustness of Sequence-to-Sequence Models with Adversarial Examples

Theme Papers

• The Unstoppable Rise of Computational Linguistics in Deep Learning
• How Can We Accelerate Progress Towards Human-like Linguistic Generalization?
• A Call for More Rigor in Unsupervised Cross-lingual Learning
• Climbing towards NLU: On Meaning, Form, and Understanding in the Age of Data
• Language (Re)modelling: Towards Embodied Language Understanding
• To Test Machine Comprehension, Start by Defining Comprehension