Supervision

📖 Thesis Supervision

🏃 Ongoing

• Jiaao Li (M.Sc.) - Title: EffiCoC: Efficient Chain-of-Code with CodeLLMs

• Fedor Splitt (B.Sc.) - Title: Counterfactual Example Generation Enhancement via Alignment-as-Preference Optimization

• Nico Tananow (M.Sc.) - Title: Are multilingual Self-Explanations faithful?

🔥 Completed

• Alexandra Iakovleva (B.Sc.) - Title: XAI in Data Privacy and Data Anonymization: A Survey
• Yilong Wang (B.Sc.) - Title: AIGC: Automatic Iterative Generation for Counterfactuals

🎓 Student Research Assistant Supervision

🚀 Ongoing

• Fedor Splitt (2025/02/01-): intent recognition in ConvXAI and counterfactual data augmentation

🎊 Completed

• Yoana Tsoneva (2025/01/16-2025/03/15): intent recognition in conversational XAI systems

🚩 Open Topics

If you are interested in any topics mentioned below, please send me your CV, transcript of record, and a short motivation letter.

1. Are Multilingual Self-Explanations Faithful?

References: Zhao and Aletras; Madsen et al. (2024); Resck et al. (2025)
Link: https://www.theses.tu-berlin.de/de/theses/0199b503-48e1-73c8-8a0b-f81bedd56a31

2. How Code-Switching impacts Explanation Faithfulness?

References: Sheth et al. (2025); Wang et al. (2025); Yoo et al. (2025); Zhao et al. (2025)
Link: https://www.theses.tu-berlin.de/de/theses/0199d213-d755-77c2-ae93-a8ce5a4a385a

3. Counterfactual Example Generation via Contrastive Activation Steering

References: Rimsky et al. (2024); Tang et al. (2025)