← People

Xinyang Geng

Source-listed: Research Scientist · Google DeepMind

  • Multimodal AI
  • Language models
  • Information retrieval

Selected work

5
  1. Gemini 1.5: Unlocking multimodal understanding across millions of tokens of contextContributor · 2024
  2. Rewarding Progress: Scaling Automated Process Verifiers for LLM ReasoningContributor · 2025
  3. OpenLLaMA: An Open Reproduction of LLaMAContributor · 2023
  4. Koala: A Dialogue Model for Academic ResearchContributor · 2023
  5. Multimodal Masked Autoencoders Learn Transferable RepresentationsFirst author · 2022