Tom Lieberum
All AI mindsTom Lieberum, full AI read
Tom Lieberum, Research Engineer, Google DeepMind, Google DeepMind (United Kingdom), ranks #424/520 on the AI Advancement Index (63.7). Known for Mechanistic interpretability research on DeepMind's team; sparse autoencoders at scale (Gemma Scope) and circuit-level analysis of transformer behavior.
Dimension read
| Dimension | Value | Standing | What a high vs low value means, and where Tom Lieberum sits |
|---|---|---|---|
| AAI AI Advancement (AAI) | 63.7 | Developing · #423/520 | Low here, lower relative influence within this elite set. ▲ high: among the very top minds advancing AI · ▼ low: lower relative influence within this elite set |
| Research influence Research influence | 58.0 | Developing · #439/520 | Low here, limited direct research influence. ▲ high: field-defining research contributions · ▼ low: limited direct research influence |
| Frontier role Frontier role | 70.0 | Moderate · #223/520 | Mid-pack. High would mean central to building today's frontier AI; low would mean removed from frontier development. ▲ high: central to building today's frontier AI · ▼ low: removed from frontier development |
| Thought leadership Thought leadership | 56.0 | Developing · #423/520 | Low here, limited public/field influence. ▲ high: shapes how the field and public think about AI · ▼ low: limited public/field influence |
| Field-building Field-building | 56.0 | Developing · #440/520 | Low here, limited field-building footprint. ▲ high: builds the field, mentorship, institutions, tools, community · ▼ low: limited field-building footprint |
| Momentum Momentum | 78.0 | Moderate · #238/520 | Mid-pack. High would mean driving AI's advancement right now; low would mean less active at the current frontier. ▲ high: driving AI's advancement right now · ▼ low: less active at the current frontier |
Strengths
- No standout dimension.
Risk factors
- A significant, well-rounded contributor to AI's advancement.
AI worldview
Ideas & positions
Tom Lieberum is a Research Engineer at Google DeepMind, focusing on AI safety and alignment, particularly through mechanistic interpretability research. He has contributed to significant projects such as sparse autoencoders at scale (Gemma Scope) and circuit-level analysis of transformer behavior. Lieberum emphasizes the importance of understanding and controlling the internal mechanisms of AI systems to ensure they behave as intended. While he has not made extensive public statements on existential risk, his work suggests a strong concern with ensuring AI systems are safe and aligned with human values. He has not taken a definitive public stance on open vs closed models or regulation, but his focus on interpretability implies a preference for transparency and accountability in AI development.
What shapes the view
Lieberum's views are shaped by his technical background in deep learning and his commitment to AI safety. His work at DeepMind, a leading institution in AI research, likely influences his emphasis on rigorous scientific methods and empirical validation. The practical challenges of ensuring AI systems are interpretable and controllable may also inform his cautious approach to AI deployment. His professional history in cutting-edge AI research suggests a pragmatic and technically grounded perspective on the potential and risks of AI.
The AI-powered future they see
Lieberum publicly predicts a future where AI systems are increasingly sophisticated and integrated into various aspects of society. He promotes the idea that through careful research and development, these systems can be made transparent and reliable, reducing the risk of unintended consequences. His work suggests a vision of AI that is both powerful and safe, with a focus on aligning AI goals with human values and ensuring that AI systems can be understood and controlled.