We meet every Friday, at 11 AM (room 3.12, Kolska office).
Propositions of papers: link
- 2023-09-22 - TBD
- 2023-09-29 - Explaining grokking through circuit efficiency - Zuzanna Kwiatkowska
- 2023-10-06 - Poisoning Language Models During Instruction Tuning - Emilia Wiśnios
- 2023-10-13 - TBD
- 2023-10-20 - TBD
- 2023-10-27 - ML in PL
- 2023-11-03 - Holiday
- 2023-11-10 - TBD
- 2023-11-17 - Direct Preference Optimization: Your Language Model is Secretly a Reward Model - Emilia Wiśnios
- 2023-11-24 - TBD - Karolina Seweryn
- 2023-12-01 - TBD
- 2023-12-08 - TBD
- 2023-12-15 - TBD
- 2023-12-22 - Holiday
- 2023-12-29 - Holiday