Publications & Open Source
Foundation Model Technical Reports
Google Scholar (opens in new tab)Listed as a contributor on Google’s model technical reports for work on AI Data metadata and provenance infrastructure.
- Jul 2025arXiv:2507.06261
Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities (opens in new tab)
Gemini Team, Google (Contributor). Covers the Gemini 2.5 model family, including Gemini 2.5 Pro and Flash.
- Mar 2025arXiv:2503.19786
Gemma 3 Technical Report (opens in new tab)
Gemma Team, Google (Contributor). Open-weights models (1B–27B) with up to 128K context, vision understanding, and multilingual support.
- Mar 2024arXiv:2403.05530
Gemini 1.5: Unlocking Multimodal Understanding Across Millions of Tokens of Context (opens in new tab)
Gemini Team, Google (Contributor). Introduces the Gemini 1.5 mixture-of-experts architecture for long-context multimodal reasoning.
- Dec 2023arXiv:2312.11805
Gemini: A Family of Highly Capable Multimodal Models (opens in new tab)
Gemini Team, Google (Contributor). Initial Gemini 1.0 release introducing Ultra, Pro, and Nano trained natively across text, image, audio, and video.
Selected Open Source
github.com/adriangoe (opens in new tab)- 2019TensorFlow.js · Magenta MusicVAE
Web app combining MusicVAE with an interactive genetic algorithm to evolve musical sequences in latent space.
- 2018Python · PyTorch
Implementation of Relation Networks (Santoro et al., NeurIPS 2017) for visual and state-based relational reasoning.
- 2017Python · PyPI · scikit-learn
Python package implementing Assumption-Free K-MC² seeding (Bachem et al., NeurIPS 2016) for fast, scikit-learn-compatible k-means initialization.
- 2016Python · Django Wagtail
Plugin for the Wagtail CMS that generates visual and field-level diffs of revised pages and posts review summaries to Slack.