Philipp Fränken

Philipp Fränken

I am a researcher at Google DeepMind working on post‑training and RL on the Gemini Team.

From 2023–2024, I was a postdoc at Stanford University working on post‑training and evals advised by Noah Goodman and Tobias Gerstenberg.

Prior to that, I completed a PhD in Psychology at The University of Edinburgh.

Twitter / Google Scholar / GitHub / Email

Selected Works

2025 JunGemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities.
2025 JanTowards System 2 Reasoning in LLMs: Learning How to Think With Meta Chain-of-Thought.
2024 AprSelf-Supervised Alignment with Mutual Information: Learning to Follow Principles without Preference Labels.
2024 MarSTaR-GATE: Teaching Language Models to Ask Clarifying Questions.
2023 OctSocial Contract AI: Aligning AI Assistants with Implicit Group Norms.
2023 JunUnderstanding Social Reasoning in Language Models with Language Models.