About Me
I am a Software Engineer at Google DeepMind, specializing in the large-scale pretraining of state-of-the-art LLMs. My research currently focuses on efficiency and performance of LLMs, but also on representation learning and neural reasoning. Feel free to browse my publications below.
Selected Publications
-
Benjamin Estermann, Roger Wattenhofer. (2025). "Reasoning Effort and Problem Complexity: A Scaling Analysis in LLMs". ICLR 2025 Workshop on Reasoning and Planning for LLMs, Singapore.[PDF]
-
Benjamin Estermann, Luca Lanzendörfer and Yannick Niedermayr and Roger Wattenhofer. (2024). "PUZZLES: A Benchmark for Neural Algorithmic Reasoning". 38th Conference on Neural Information Processing Systems (NeurIPS), Vancouver, Canada.
-
Benjamin Estermann, Roger Wattenhofer. (2023). "DAVA: Disentangling Adversarial Variational Autoencoder". 11th International Conference on Learning Representations (ICLR), Kigali, Rwanda.
-
Benjamin Estermann*, Markus Marks* and Mehmet Fatih Yanik. (2020). "Robust Disentanglement of a Few Factors at a Time using rPU-VAE". 34th Conference on Neural Information Processing Systems (NeurIPS), Vancouver, Canada.
For a complete list, please see my Google Scholar profile.