Alicia Vail Parrish
Source-listed: Senior research scientist · Google DeepMind
Record snapshot: Sep 3, 2026. Coverage and affiliations may be incomplete or historical.
Selected work
5- Gemma 2: Improving Open Language Models at a Practical SizeSaved credit: Contributor · 2024Explore people connected to this work →
- Adversarial Nibbler: An Open Red-Teaming Method for Identifying Diverse Harms in Text-to-Image GenerationSaved credit: Contributor · 2024Explore people connected to this work →
- DICES Dataset: Diversity in Conversational AI Evaluation for SafetySaved credit: Contributor · 2023Explore people connected to this work →
- BBQ: A Hand-Built Bias Benchmark for Question AnsweringSaved credit: Contributor · 2022Explore people connected to this work →
- Going PLACES: Participatory Localized Red Teaming for Text-to-Image Safety in the Global SouthSaved credit: Contributor · 2026Explore people connected to this work →