Yifan Zhang · University of Ostrava — LLM work from generative UI design agents to power-system transient stability
2
papers here
2
reviewed by Evolve
1
first-author
0
corresponding
—
works · OpenAlex
0
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-10-15) 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—
Output metrics are not yet available for this identity, so the axis is left unscored. It would measure total publication volume, venue mix, and consistency of scholarly output over time; the current record is too sparse to compute these reliably.
Impact—
Impact data is insufficient for scoring: the recorded h-index and i10-index are both 0 and no citation series is available. This axis would measure citations per year, field-normalized influence, and uptake of the two cataloged papers.
Focus0/100
The focus axis scores 0, reflecting extreme topical dispersion in the underlying record. Evolve's field distribution splits evenly at 0.5/0.5 between power-systems machine learning and HCI/generative AI, while the OpenAlex topic list scatters across advanced graph theory, combinatorial mathematics, nonlinear PDEs, speech recognition, and curvature flows. A split this broad — including topics that do not match the cataloged papers — suggests the identity record likely mixes several researchers named Yifan Zhang rather than one concentrated agenda.
Collaboration44/100
Collaboration scores 44, a mid-range value built on 8 distinct coauthors in the record. The frequent-coauthor list — Chao Shen, Hongwei Zhen, Junyan Shao, Zhenghao Yang, Mingyang Sun, Nghi D. Q. Bui, Georgios Evangelopoulos, and Arnaud Benard — suggests small team-based arXiv work rather than large consortium science. The two cataloged papers fit this pattern: one first-authored, one middle-authored, both multi-author. Institutional spread behind each name cannot be verified from the available data.
Activity75/100
Activity scores 75 on strong recent volume: the record logs 12 works in the past three years, and counts-by-year attributes all 12 to 2026. Both cataloged papers appeared on arXiv in 2026. The concentration cuts both ways — it marks a productive current phase, but with no earlier years in the series the trajectory rests on a single snapshot year, consistent with an early-stage or newly activated profile.
Reproducibility20/100
Reproducibility scores 20, with a reproducibility mean of 20 across the 2 papers Evolve has reviewed. Both carry a low repro-readiness rating: the vibe-design-agents paper grounds quality claims in an LLM judge, and LLaTSA reports NRMSE improving from 0.71 to 0.31 against narrow baselines on synthetic grids without real-system validation. Neither currently offers the evidence base for quick independent verification.
Research summary
Yifan Zhang is listed with the University of Ostrava (Czech Republic) in an OpenAlex-derived record; this profile is name-resolved from public bibliographic data, with no ORCID to anchor identity. Evolve's catalog holds two 2026 arXiv papers. The first, "Enabling Creative Exploration for Vibe Design Agents" (first author), is reviewed by Evolve as a Google study adding a Verbalized-Sampling spec-selection stage to UI design agents: exploration broadens measurably, though quality evidence rests on an LLM judge. The second, "LLaTSA" (middle author), claims the first LLM-aligned general-purpose transient stability analysis, reporting NRMSE improving from 0.71 to 0.31 versus Uni-TSA on synthetic grids — with narrow baselines and no real-system validation. The record logs 12 works in the past three years, all attributed to 2026, alongside 8 distinct coauthors including Chao Shen, Mingyang Sun, and Nghi D. Q. Bui. Dimension scores put activity high (75), collaboration mid-range (44), and reproducibility low (20 across the 2 reviewed papers), while output and impact are left unscored because citation data is missing — the recorded h-index and i10-index are both 0. A notable caveat: the OpenAlex topic profile (graph theory, combinatorial mathematics, nonlinear PDEs, speech recognition, geometric analysis) does not match the cataloged machine-learning papers, so part of the metric record may belong to other researchers sharing this common name.
Collaboration pattern
The record lists 8 distinct coauthors and scores collaboration at 44, mid-range. Frequent coauthors include Chao Shen, Hongwei Zhen, Junyan Shao, Zhenghao Yang, Mingyang Sun, Nghi D. Q. Bui, Georgios Evangelopoulos, and Arnaud Benard — a mix suggesting small, team-based arXiv projects rather than large consortia. The two cataloged papers fit this: one first authorship, one middle authorship, both multi-author. Which coauthors sit on which paper cannot be determined from the available data.
Research keywords
Highlights
Verbalized Sampling broadens exploration for vibe design agents
LLaTSA: first LLM-aligned general-purpose transient stability analysis
Low reproducibility readiness across the reviewed output
Both Evolve-reviewed papers carry a low repro-readiness rating, giving a reproducibility mean of 20 over 2 papers — quality claims lean on an LLM judge in one and on synthetic-grid benchmarks with narrow baselines in the other.
Timeline
- 2026Co-authored (middle position) LLaTSA on arXiv, claimed as the first LLM-aligned general-purpose transient stability analysis, with NRMSE 0.71→0.31 versus Uni-TSA on synthetic grids.
- 2026First-authored "Enabling Creative Exploration for Vibe Design Agents" on arXiv, adding a Verbalized-Sampling spec-selection stage to UI design agents.
- 2026The OpenAlex record attributes all 12 recent works to this single year, with 8 distinct coauthors — activity scores 75.
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 come from an OpenAlex identity record only; no ORCID is provided, and "Yifan Zhang" is a common name, so this profile is name-resolved from public bibliographic data.
- The OpenAlex topic profile (advanced graph theory, combinatorial mathematics, nonlinear PDEs, speech recognition, geometric analysis and curvature flows) does not match the two cataloged machine-learning papers, so part of the metric record may belong to different researchers sharing the name.
- Output and impact axes are unscored: no reliable citation series is available, and the recorded h-index and i10-index are both 0.
- Reproducibility rests on only the 2 Evolve-reviewed papers, both low-readiness arXiv preprints; no peer-reviewed venues are recorded in the available data.