PeopleofSuperintelligence

Researcher

Jenna Sparks

Jenna Sparks works with AI teams on scientific reading, search and open vision-language models. She coauthored an ecology paper in 2021, contributed to Semantic Reader's skimming project in 2023 and worked as a data engineer on UC Berkeley's Very Intelligent Portfolio capstone in 2024. In 2024, she coauthored CHIME, Context24 and Molmo with its human-collected PixMo datasets; the Molmo paper is an oral paper and award candidate at CVPR 2025. In February 2026, she coauthored OpenScholar for Nature with an Ai2 affiliation in Seattle; its benchmark covers computer science, physics, neuroscience and biomedicine.

Timeline

  1. 2021

    Island intertidal recovery

    She has a background in environmental studies and biology from UC Santa Cruz. In 2021, she coauthors a Scientific Reports study with UC Santa Cruz's Conservation Action Lab and Oikonos Ecosystem Knowledge on intertidal recovery after rat removal. The paper reports that the rocky intertidal community on Hawadax Island returns to resemble communities on rat-free islands.

  2. October 2023

    Semantic Reader skimming

    In October 2023, she contributes to Semantic Reader's skimming project at Ai2, a free scientific PDF reader with AI features. Its skimming feature uses automatically generated highlights to help researchers scan scientific papers.

  3. 2024

    Very Intelligent Portfolio

    In 2024, she works as a data engineer with Rachel Gao, Shuo Wang and Ray Cao on UC Berkeley's MIDS capstone Very Intelligent Portfolio. She is also on Ai2's Data Science Team. The team's open-source machine-learning tool optimizes portfolio weights and recommends hedging strategies for employees with concentrated stock-compensation risk.

  4. July 2024

    CHIME literature hierarchies

    In July 2024, she coauthors CHIME with colleagues at the Allen Institute for AI. The paper explores LLM-created hierarchies for organizing scientific studies to support literature reviews; its expert-curated biomedical dataset contains 2,174 generated hierarchies covering 472 topics.

  5. August 2024

    Context24 scientific claims

    In August 2024, she coauthors the Context24 overview for the Fourth Workshop on Scholarly Document Processing in Bangkok. The shared-task dataset pairs 585 scientific claims with gold annotations for supporting figures and tables and text snippets of methodological details.

  6. 2024–2025

    Molmo and PixMo at CVPR

    In September 2024, she coauthors Molmo, an open family of vision-language models from the Allen Institute for AI (Ai2). Its PixMo datasets include detailed captions collected from human annotators using speech-based descriptions, free-form image questions and answers, and 2D pointing data, all gathered without external vision-language models. At CVPR in 2025, the paper is an oral paper and an award candidate; its authors call the quality of their new datasets the most critical factor in the models' success.

  7. February 2026

    OpenScholar and ScholarQABench

    In February 2026, she coauthors the Nature paper OpenScholar with an Ai2 affiliation in Seattle; it is a retrieval-augmented language model that answers scientific queries using 45 million open-access papers. Its ScholarQABench benchmark contains 2,967 expert-written queries across computer science, physics, neuroscience and biomedicine, and OpenScholar reaches human-expert levels of citation accuracy.

Career

  1. NowAllen Institute for AI (Ai2)

Education

  1. University of California, Santa Cruz

From the biography.

  1. The company share of AI researchers’ top papers was 31% in 2020–2024 and 8% in 1970–1974Includes Allen Institute for AI (Ai2). We estimate where the people shaping AI worked when their most-cited papers appeared: the company share was 31% in 2020–2024 and 8% in 1970–1974, and Google led all organizations in 2020–2024.
  2. Companies supply the most prior-work hires at 7 of the 9 frontier AI labsIncludes Allen Institute for AI (Ai2). Our estimates of the organizations where 550 people worked directly before joining 9 frontier AI labs across 620 career transitions, comparing the main feeder organizations per lab.
  3. 72% of AI researchers with top works connect into a single 2,592-person co-author groupIncludes Allen Institute for AI (Ai2). We map the co-author network across 3,593 people shaping AI and 14,977 most-cited papers: 2,592 (72%) form one connected group with a median distance of 5 steps, led in 2-step reach by Yoshua Bengio (570) and in shortest-path bridging by Yoshua Bengio (6.7%).
  4. Direct Form 990 pay in the latest returns of 39 nonprofits with AI people on this site is highest at $2.1M at the Allen Institute for AI (Ai2) (2024)Includes Allen Institute for AI (Ai2). IRS Form 990 compensation across 39 nonprofits with AI people on this site from 2012 to 2025: in the latest returns, Mozilla reports $6.1M including related-company pay (2024) and the Allen Institute for AI (Ai2) leads direct nonprofit pay at $2.1M (2024).

Sources and credits

Photo: We have no published photo of Jenna Sparks yet. Send us a free-licensed photo.

  1. Indirect effects of invasive rat removal result in recovery of island rocky intertidal community structure | Scientific Reports nature.com. Checked 21 October 2026.
  2. VIP - Very Intelligent Portfolio very-intelligent-portfolio.github.io. Checked 21 October 2026.
  3. Case study: Iterative design for skimming support | Ai2 allenai.org. Checked 21 October 2026.
  4. Very Intelligent Portfolio | UC Berkeley School of Information ischool.berkeley.edu. Checked 21 October 2026.
  5. CHIME: LLM-Assisted Hierarchical Organization of Scientific Studies for Literature Review Support arxiv.org. Checked 21 October 2026.
  6. Overview of the Context24 Shared Task on Contextualizing Scientific Claims - ACL Anthology aclanthology.org. Checked 21 October 2026.
  7. Molmo and PixMo: Open Weights and Open Data for State-of-the-Art Multimodal Models arxiv.org. Checked 21 October 2026.
  8. CVPR Oral Molmo and PixMo: Open Weights and Open Data for State-of-the-Art Vision-Language Models cvpr.thecvf.com. Checked 21 October 2026.
  9. Synthesizing scientific literature with retrieval-augmented language models | Nature nature.com. Checked 21 October 2026.
  10. In a study, AI model OpenScholar synthesizes scientific research and cites sources as accurately as human experts | UW News washington.edu. Checked 21 October 2026.