Responsible Generative Artificial Intelligence

Gizem Gezici

The exam will consist of a project and its presentation with a discussion.

This course has no prerequisites.

The course is designed for Master's students of the Scuola Normale Superiore (SNS) and PhD students enrolled in the National Ph.D. of Artificial Intelligence (University of Pisa)

and the Computational methods and mathematical models for science and finance (SNS). PhD students from other SNS doctoral programs with an interest in the topics covered by

the course are also encouraged to enroll.

The course is organized as follows in three modules: 

Module I – The Evolution of Generative AI

Evolution of NLP: from rule-based systems to Large Language Models (LLMs).

Module II – Technical Foundations of Generative AI

Transformer architecture, transfer learning, and large-scale computation.

Module III – Responsible Generative AI

Ethical, social, and technical challenges, including bias, hallucinations, and interpretability.

Evaluation and mitigation strategies.

Seminars, case studies, and hands-on sessions.

Reference bibliography

1. Vaswani, Ashish, et al. "Attention is all you need." Advances in neural information processing systems 30 (2017).
2. Paaß, G., & Giesselbach, S. (2023). Foundation models for natural language processing: Pre-trained language models integrating media (p. 436). Springer Nature.
3. Devlin, Jacob, et al. "Bert: Pre-training of deep bidirectional transformers for language understanding." Proceedings of the 2019 conference of the North American chapter of the association for computational linguistics: human language technologies, volume 1 (long and short papers). 2019.
4. Radford, Alec, et al. "Improving language understanding by generative pre-training." (2018).
5. Brown, Tom, et al. "Language models are few-shot learners." Advances in neural information processing systems 33 (2020): 1877-1901.
6. Wei, Jason, et al. "Chain-of-thought prompting elicits reasoning in large language models." Advances in neural information processing systems 35 (2022): 24824-24837.
7. Kojima, Takeshi, et al. "Large language models are zero-shot reasoners." Advances in neural information processing systems 35 (2022): 22199-22213.
8. Ouyang, Long, et al. "Training language models to follow instructions with human feedback." Advances in neural information processing systems 35 (2022): 27730-27744.
9. Karpukhin, Vladimir, et al. "Dense Passage Retrieval for Open-Domain Question Answering." EMNLP (1). 2020.
10. Dettmers, Tim, et al. "Qlora: Efficient finetuning of quantized llms." Advances in neural information processing systems 36 (2023): 10088-10115.
11. Gallegos, Isabel O., et al. "Bias and fairness in large language models: A survey." Computational Linguistics 50.3 (2024): 1097-1179.
12. Shuster, Kurt, et al. "Retrieval augmentation reduces hallucination in conversation." Findings of the Association for Computational Linguistics: EMNLP. 2021.
13. Goodfellow, Ian, et al. "Generative adversarial networks." Communications of the ACM 63.11 (2020): 139-144.
14. Bommasani, Rishi, et al. "On the opportunities and risks of foundation models." arXiv preprint arXiv:2108.07258 (2021).
15. EMNLP 2024 Tutorial: Language Agents: Foundations, Prospects, and Risks https://language-agent-tutorial.github.io/