Carnegie Mellon researcher bridging software engineering, explainable AI, and open benchmark design for LLM evaluation
2
papers here
1
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
11
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: the input data required to compute publication output is unavailable in the current snapshot, so this axis is left unscored; it would measure total publication volume and productivity consistency once that data exists.
Impact—
Not yet scored: the citation-based inputs for this axis are insufficient in the current data, so impact is left unscored; with an h-index of 11 recorded, the axis would measure citation uptake and scholarly influence of the work.
Focus100/100
The perfect focus score of 100 mirrors a tightly clustered topic profile: Software Engineering Research (33 works) and Software Engineering Techniques and Practices (28) dominate, joined by Explainable AI (19), Teaching and Learning Programming (13), and Software Testing and Debugging Techniques (12), with only one clinical-AI paper in the field distribution — essentially one coherent software-and-XAI agenda rather than scattered interests.
Collaboration41/100
The collaboration score of 41 rests on a small recent network: 7 distinct coauthors in the three-year window, the lowest of the scored axes. The listed frequent coauthors — Christine Park, Tim Dettmers, Nishant Balepur, Connor Baumler, Eunsol Choi, Rachel Rudinger, Jordan Boyd-Graber — span NLP and ML groups, and both catalog papers are middle-authored team efforts.
Activity100/100
The perfect activity score tracks an accelerating output curve: yearly works grew from 3 in 2019 and 2 in 2020 to 4 in 2022, 7 in 2023, 15 in 2025, and 30 in 2026, with 51 works in the most recent three years — the steepest ramp in the record and clear evidence of sustained, intensifying research engagement.
Reproducibility100/100
A mean reproducibility score of 100 is recorded, but the base is thin: only 1 catalog paper has been assessed — Synthetic Hospital (2026), which Evolve rates high on reproducibility readiness for its open, traceable, physician-validated design. Treat the perfect score as an early signal on a single-paper sample, not an established track record.
Research summary
Valerie Chen is a researcher at Carnegie Mellon University whose OpenAlex-resolved profile sits at the intersection of software engineering and explainable AI. The topic profile is tightly concentrated: Software Engineering Research (33 works) and Software Engineering Techniques and Practices (28) lead, followed by Explainable AI (19), Teaching and Learning Programming (13), and Software Testing and Debugging Techniques (12). Activity is the standout signal: yearly works climb from 3 in 2019 and 2 in 2020 to 7 in 2023, 15 in 2025, and 30 in 2026, with 51 works in the most recent three years — the basis for a perfect activity score. The citation footprint is still early-stage (h-index 11, i10-index 11), and the output and impact radar axes remain unscored for lack of sufficient input data. Recent work extends the benchmark-and-evaluation agenda into new territory: Synthetic Hospital (2026) is an open, physician-validated synthetic longitudinal EHR benchmark built on FHIR R4 with traceable labels and a knowledge graph, rated high on reproducibility readiness — the single assessed paper behind a reproducibility mean of 100 — while VibeJam (2026) builds a user-study platform for web development with AI agents. Collaboration metrics show 7 distinct recent coauthors (collaboration score 41), with named frequent coauthors spanning NLP and machine learning groups. Both catalog papers are middle-authored, non-corresponding contributions to larger team efforts.
Collaboration pattern
Collaboration data points to a small but active recent network: 7 distinct coauthors in the three-year window, and both catalog papers are middle-position, non-corresponding team efforts. The frequent-coauthor list — Christine Park, Tim Dettmers, Nishant Balepur, Connor Baumler, Eunsol Choi, Rachel Rudinger, and Jordan Boyd-Graber — connects this profile to NLP and ML groups, consistent with a researcher who plugs into larger benchmark and evaluation projects rather than leading solo lines.
Research keywords
Highlights
Synthetic Hospital: an open, physician-validated EHR benchmark
VibeJam: a user-study platform for agent-driven web development
The 2026 arXiv paper VibeJam builds a platform for studying how developers build web software with AI agents, extending the benchmark-and-evaluation agenda toward human-centered agent evaluation; Evolve has not yet published a verdict, keywords, field, or reproducibility rating for it.
Timeline
- 2019Earliest year in the publication record: 3 works logged in the OpenAlex counts-by-year data.
- 2023Productivity starts to rise: 7 publications in the year, nearly double the 4 of 2022.
- 2025Acceleration phase: 15 publications in the year; 51 works span the most recent three years.
- 2026Busiest year on record with 30 publications; two arXiv papers — the Synthetic Hospital clinical-AI benchmark and the VibeJam agent user-study platform — enter Evolve's review catalog.
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 metrics (h-index, i10-index, counts by year, topics, coauthors) come from OpenAlex; the profile for the name 'Valerie Chen' is name-resolved from public bibliographic data, so common-name ambiguity is possible.
- The output and impact dimensions are unscored because the inputs for those axes are insufficient in the current snapshot; no citation counts beyond h-index 11 and i10-index 11 were provided.
- The perfect reproducibility score rests on a single assessed paper (reproPapers = 1) and should not be read as an established track record.
- VibeJam appears in the catalog without a verdict, keywords, field, reproducibility rating, or Evolve slug, limiting what can be claimed about it.