Scholar
Phillip Lippe
Google Scholar ID: 69hFZp4AAAAJ
Google DeepMind
Citations & Impact
Citations
1,638
H-index
15
i10-index
19
Publications
20
Co-authors
15
list available
Publications
Resume (English only)
Academic Achievements
- Paper 'How to Train Neural Field Representations - A Comprehensive Study and Benchmark' accepted at CVPR 2024.
- Paper 'Towards the Reusability and Compositionality of Causal Representations' accepted as oral at CLeaR 2024.
- Two papers accepted at NeurIPS 2023: 'PDE-Refiner: Achieving Accurate Long Rollouts with Neural PDE Solvers' (Spotlight) and 'Rotating Features for Object Discovery' (Oral).
- Paper 'Modeling Accurate Long Rollouts with Temporal Neural PDE Solvers' accepted as Oral at Frontiers4LCD workshop, ICML 2023.
- Paper 'BISCUIT: Causal Representation Learning from Binary Interactions' accepted as Spotlight at UAI 2023.
- Two posters presented at NeurIPS 2023 Workshop on Causal Representation Learning: 'Hierarchical Causal Representation Learning' and 'Towards the Reusability and Compositionality of Causal Representations'.
- Led research effort that brought Imagen 4 back to #1 on LMArena (July 2025).
- Gemini 2.5 Flash Image (aka Nano Banana) launched in August 2025 with a 171 Elo lead on LMArena.
- Co-organized CVPR 2024 Workshop on Causal and Object-Centric Representations for Robotics.
- Serving as Workflow Chair for UAI Conference 2025.
- Lecturer at ELLIS European Summer School on Information Retrieval (ESSIR 2024), teaching recent advances in Transformers, LLMs, and scaling.
- Released 10 new tutorial notebooks in the UvA-DL series implementing data, pipeline, and tensor parallelism from scratch in JAX+Flax.
