Oxford researcher in agentic medical AI — recursive self-improvement, clinical safety, and AI persuadability
2
papers here
1
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
0
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-10-17) and cover the author's entire public record, not just the papers listed here. Reproducibility signal comes from Evolve's own reviews.
Research profile
Output—
Not scored yet: OpenAlex provides no reliable publication-volume signal for this profile, and the Evolve catalog holds only two 2026 arXiv preprints. This axis would measure total works, venue types, and authorship roles across the full publication record.
Impact—
Not scored yet: citation impact cannot be computed — h-index and i10-index are recorded as 0 and both known works are recent 2026 arXiv preprints. This axis would measure citations, downstream adoption, and scholarly influence once citations accumulate.
Focus100/100
Scores 100/100 for focus. The topic distribution is tightly concentrated: Artificial Intelligence in Healthcare and Education accounts for 3 of the recorded works, with Ethics and Social Impacts of AI, Biomedical Text Mining and Ontologies, Scientific Computing and Data Management, and Explainable AI each appearing once — all orbiting the same medical-AI theme. The single classified field entry, medical artificial intelligence (agentic systems), matches the MedRSI paper's framing.
Collaboration39/100
Scores 39/100. The record shows 6 distinct coauthors, with Junde Wu, Jiayuan Zhu, Minghao Hu, Jiazhen Pan (listed twice in the source data), and Minhao Hu appearing most frequently. Both catalog papers place Liu in a middle author position without corresponding-author status, consistent with contributing within a larger Oxford-based medical AI team rather than leading an independent group.
Activity40/100
Scores 40/100. Activity is concentrated entirely in 2026: the OpenAlex counts-by-year record shows 3 works, and recent3yWorks is likewise 3, meaning no earlier output appears under this profile. This pattern is consistent with an emerging researcher whose visible record has just begun, rather than a dip late in an established career; the score typically rises as the record accumulates.
Reproducibility55/100
Scores 55/100, computed from 1 assessed paper (reproMean 55). The only reviewed work, MedRSI, carries a medium reproducibility readiness: Evolve's verdict notes strong benchmark gains on glaucoma diagnosis and echocardiography, but flags LLM-judged clinical cost and reliance on private data as limits on independent verification. One assessed paper makes this a provisional signal.
Research summary
Fenglin Liu is a researcher at the University of Oxford working at the intersection of medical artificial intelligence and agentic LLM systems. The Evolve catalog lists two 2026 arXiv papers, both with Liu in a middle author position and neither as corresponding author. The first, MedRSI: Recursive Self-Improvement for Medical Agents via Clinically Aligned Self-Evolution, claims the first medical recursive self-improvement framework, reporting strong benchmark gains in glaucoma diagnosis and echocardiography; Evolve's review notes that LLM-judged clinical costs and private data limit independent verification, and rates reproducibility readiness as medium. The second, AI Can Be Easily Persuaded in Clinical Decision Making, addresses the persuadability of AI in clinical decision making; Evolve has not yet published a verdict for it. OpenAlex records (identity resolved from public bibliographic data) show 3 works in 2026 and 3 works over the last three years, with topics led by Artificial Intelligence in Healthcare and Education (3), followed by Ethics and Social Impacts of AI, Biomedical Text Mining and Ontologies, Scientific Computing and Data Management, and Explainable AI (1 each). Six distinct coauthors appear, most frequently Junde Wu, Jiayuan Zhu, Minghao Hu, Jiazhen Pan, and Minhao Hu. Citation indicators (h-index and i10-index both 0) reflect a just-starting record rather than an established citation footprint.
Collaboration pattern
Fenglin Liu's record lists 6 distinct coauthors, with Junde Wu, Jiayuan Zhu, Minghao Hu, Jiazhen Pan, and Minhao Hu recurring most frequently — an Oxford-centered collaboration cluster. Both catalog papers place Liu as a middle author without corresponding-author status, suggesting contributions within team-led projects rather than independent leadership. The source data lists Minghao Hu and Minhao Hu as separate names; no ORCID is available to disambiguate.
Research keywords
Highlights
MedRSI: recursive self-improvement for medical agents
AI Can Be Easily Persuaded in Clinical Decision Making
Second 2026 arXiv paper in the Evolve catalog, addressing the persuadability of AI systems in clinical decision making. Evolve has not yet published a review verdict, keywords, or reproducibility rating for this paper, so its assessment is pending.
Highly concentrated early record in healthcare AI
OpenAlex records 3 works, all in 2026, with the topic Artificial Intelligence in Healthcare and Education counted 3 times, and AI ethics and social impacts, biomedical text mining, scientific computing and data management, and XAI counted once each — a narrow, coherent early agenda.
Timeline
- 2026Posts MedRSI on arXiv, a recursive self-improvement framework for medical agents (middle author, not corresponding).
- 2026Posts “AI Can Be Easily Persuaded in Clinical Decision Making” on arXiv (middle author, not corresponding).
- 2026OpenAlex counts 3 works in 2026 — the entire visible record to date, concentrated in healthcare AI topics.
Papers in the Evolve catalog
Frequent co-authors
About this profile
This profile is AI-structured by Evolve.Pub from public bibliographic data (OpenAlex) and Evolve's own paper reviews. It is not written or endorsed by the researcher unless the profile is claimed. Radar dimensions are computed from real bibliometrics with fixed formulas — raw figures are shown alongside; missing dimensions mean insufficient data, never zero. Errors? Claim the profile to correct or hide it.
- Bibliometric indicators (works, topics, coauthors, h-index, i10-index) come from OpenAlex only; identity is name-resolved from public bibliographic data.
- No ORCID is linked in the source data, which limits author disambiguation; “Fenglin Liu” is a relatively common name.
- Only two papers by this researcher are in the Evolve catalog, and only MedRSI has a review verdict, keywords, and a reproducibility rating.
- Output and impact dimensions are unscored: h-index and i10-index are 0 in the source data and both works are recent arXiv preprints.