Last author of 2026 arXiv work on autonomous LLM research agents, LLM hypothesis discovery, and human-aware dialogue.
3
papers here
3
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
—
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-09-20) 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—
This axis would measure total publication volume and productivity over time, but this snapshot contains only the three reviewed 2026 arXiv papers with no full publication count, so the dimension is left unscored rather than estimated.
Impact—
Impact would measure citations and downstream influence of this researcher's work. No citation counts, h-index, or venue acceptance data are present in the snapshot, so the axis is reported as missing rather than scored.
Focus0/100
Focus scores 0/100 because the field distribution is perfectly even: the three reviewed papers fall one each into autonomous LLM research agents, natural language processing, and AI-for-science hypothesis discovery (0.333 each). The work spreads across three adjacent LLM-agent subfields rather than concentrating in one.
Collaboration73/100
Collaboration scores 73/100, backed by 37 distinct coauthors across the three reviewed papers and a recurring core that includes Zhiting Hu, Xinle Yu, Fan Bai, Jefferson Chen, and Pratibha Jagannatha. Sitting last author on all three papers fits a senior, group-oversight role within a broad collaboration network.
Activity—
Activity would measure recent publishing cadence — papers per year and gaps between outputs. The snapshot only fixes three arXiv releases in 2026 and provides no earlier or later records, so cadence cannot be computed and the axis stays unscored.
Reproducibility32/100
Reproducibility averages 32/100 over three assessed papers (reproPapers = 3, reproMean = 32). PrimeScientist is rated medium, while Mind2Dialogue and HypoEvolve are rated low — matching Evolve's verdicts that headline gains rest on single runs, closed teacher models, and unresolved baseline gaps.
Research summary
Zhen Wang's Evolve profile is built from three 2026 arXiv preprints on which this researcher appears as last author — the conventional senior-author slot — though none lists them as corresponding author. The three papers spread evenly across three fields (0.333 each in the field distribution): autonomous LLM research agents, natural language processing, and AI-for-science hypothesis discovery, which yields the low focus score of 0/100. PrimeScientist applies budget-adaptive Monte Carlo tree search over executable plan trees to allocate autonomous-research effort; HypoEvolve coordinates LLM agents with a generational genetic algorithm for drug-repurposing hypotheses; Mind2Dialogue trains human-aware assistants from an oracle that sees simulated user mental states. Collaboration is the strongest measured axis (73/100), with 37 distinct coauthors across the reviewed papers and recurring collaborators including Zhiting Hu, Xinle Yu, and Fan Bai. Reproducibility averages 32/100 over three assessed papers, with per-paper readiness rated medium, low, and low — mirroring Evolve's notes on single-run results and closed teacher models. Publication counts, citations, and yearly activity are absent from this snapshot and remain unscored.
Collaboration pattern
The 73/100 collaboration score rests on 37 distinct coauthors across the three 2026 papers, with a recurring core that includes Zhiting Hu, Xinle Yu, Fan Bai, Jefferson Chen, JungHo Kong, and Pratibha Jagannatha. All three papers place Zhen Wang in the last (senior) author slot without corresponding-author designation, suggesting group oversight within a large, rotating team rather than a fixed two-person partnership.
Research keywords
Highlights
Budget-adaptive MCTS allocates autonomous-research effort (PrimeScientist)
Genetic algorithms coordinate LLM agents for hypothesis discovery (HypoEvolve)
Simulated user mental states train human-aware assistants (Mind2Dialogue)
Timeline
- 2026PrimeScientist: strategic allocation of research effort in autonomous research (arXiv, last author).
- 2026Mind2Dialogue: training human-aware language models by simulating user mental states (arXiv, last author).
- 2026HypoEvolve: genetic algorithms enable multi-agent LLMs to discover scientific hypotheses (arXiv, last author).
- 2026All three reviewed works appear in the same year, spanning autonomous agents, NLP, and AI-for-science.
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.
- Bibliometrics derive from OpenAlex-based public bibliographic data only; institution, country, and ORCID are absent from this snapshot.
- 'Zhen Wang' is a common name: this profile is name-resolved from public bibliographic records, and a same-name string even appears in the supplied coauthor list, so entity resolution may carry noise.
- Output, impact, and activity are unscored because citation counts, h-index, and per-year publication histories were not available in the input data.
- All three reviewed papers are 2026 arXiv preprints; peer-review venue status is not reflected in this data.