Week 2: Feb. 1

Topic: Overview of Interpretability, Explainability, and Robustness 

Week 4: Feb. 15

Topic: Mitigating Bias

Reading:

• BLIND: Bias Removal With No Demographics 

• Improving Bias Mitigation through Bias Experts in Natural Language Understanding

Week 6: Feb. 29

Topic: Privacy

Reading:

• What Does it Mean for a Language Model to Preserve Privacy? 

• Scalable Extraction of Training Data from (Production) Language Models

Week 8: Mar. 14

Topic: Explainability — Feature Attribution

Reading:

• EXPLAIN, EDIT, GENERATE: Rationale-Sensitive Counterfactual Data Augmentation for Multi-hop Fact Verification

• XMD: An End-to-End Framework for Interactive Explanation-Based Debugging of NLP Models 

Week 10: Mar. 28

Topic: Explainability — Free-Text

Reading:

• The Unreliability of Explanations in Few-shot Prompting for Textual Reasoning

• On Measuring Faithfulness of Natural Language Explanations 

Week 12: Apr. 11

Topic: Robustness — Distribution Shift

Reading:

• Robust Prompt Optimization for Large Language Models Against Distribution Shifts 

• Revisiting Out-of-distribution Robustness in NLP: Benchmark, Analysis, and LLMs Evaluations

Week 14: Apr. 26

Topic: Guest Lecture