Tufa Talks episode 3 - Valentina Pyatkin (Allen Institute for AI)

​Curated and hosted by Matteo Saponati in collaboration with Tufa Labs, an independent research group working on fundamental AI research.

​Join us! Our goal is to share the latest ML research and foster open discussion and collaboration among ML researchers in Zürich.

​Speaker: Valentina Pyatkin (Allen Institute for AI)
https://valentinapy.github.io/

​Valentina Pyatkin is a researcher and academic, currently a postdoctoral scientist at the Allen Institute for AI and the University of Washington, and an affiliate of the ETH AI Center. Her work sits at the intersection of open science and language model development, with a specific focus on post-training pipelines, verifiable instruction following, and rigorous, human-centered evaluation. A core contributor to projects like Tulu, Open-Instruct, and OLMo, she is driven by the conviction that the future of generative AI depends on making models that are contextually robust, transparent, and built on verifiable scientific foundations. Beyond her research, she is deeply involved in the community, co-organizing workshops on social responsibility in AI and mentoring students through the Swiss AI Initiative. The thread running through her career is a commitment to advancing the "open science" of LLMs, ensuring that the development of powerful AI remains collaborative, reproducible, and aligned with human values.

​Agenda:

​📅 Date: June 29
🕕 Time: 6:00–8:00 PM
📍 Place: Tufa Labs

​17:45 – Doors open
18:00 – Talk
18:30 – Open Q&A
19:00 – Pizza, drinks, networking

​Please note: doors open at 17:45 and close promptly at 18:00. Please arrive within that window.

​We'll have pizzas and drinks after the talk, hope you can stay!

​Spots are limited, and registration requires approval. If selected, you will be notified by Friday before the event.