LLM agent memory and evaluation statistics researcher at UT Dallas — 24 OpenAlex-indexed works, h-index 5
2
papers here
0
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
5
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-09-27) 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—
No output score is computed yet. This axis measures publication volume and venue strength over time; the snapshot provides per-year work counts (24 works since 2014, peaking at 7 in 2023) but no comparable composite output measure.
Impact—
Impact is not yet scored. The only signals present are an h-index of 5 and an i10-index of 3; without per-paper citation counts or field-normalized comparisons, a meaningful impact score cannot be derived.
Focus—
Focus is not scored. The snapshot lists five topic clusters led by Multi-Agent Systems and Negotiation (7 works), but topic data lacks the temporal depth and field classification needed for a concentration score.
Collaboration44/100
The collaboration score is 44. The OpenAlex snapshot lists 8 distinct coauthors, with Yezhou Cheng, Runjia Du, Zeming Liu, Qibai Chen, Hang Lyu, Bojun Lin, and Yilan Wei recurring most often. Across 24 indexed works this averages roughly three coauthors per work, and both 2026 Evolve-catalog papers place Zeng in a middle author position rather than first or corresponding, indicating a steady but compact collaborative footprint.
Activity57/100
The activity score is 57. Counts by year show continuous presence from 2014 (1 work) through 2025 (3 works), peaking at 7 works in 2023, with gaps in 2015–2016 and 2020. Six works fall within the last three years, and two further 2026 arXiv preprints sit in the Evolve catalog — an active, sustained research pace.
Reproducibility—
Reproducibility is not scored at profile level. Evolve's reviews rate the two 2026 catalog papers' repro-readiness as medium and high, but no aggregate artifact-availability or replication data exists for the wider body of work.
Research summary
Yankai Zeng is a researcher at The University of Texas at Dallas working across large language model (LLM) agents, multi-agent systems and negotiation, topic modeling, and logic, reasoning, and knowledge. An OpenAlex-resolved record (ORCID 0009-0007-6817-9747) lists 24 indexed works spanning 2014–2025, with an h-index of 5 and an i10-index of 3; output peaked at 7 works in 2023, with 6 works in the last three years. Two 2026 arXiv preprints in the Evolve catalog carry middle author positions. 'Scope Before You Persist' proposes scoped retrieval to prevent cross-family interference in persistent agent memory; Evolve's review credits solid paired evidence while noting a single templated benchmark and a single model, with medium repro-readiness. 'Sharp Limits for Honest Uncertainty' proves sharp replication limits for hard-budget repeated LLM evaluation using matching designs and equal-budget evidence; practical gains are regime-dependent, and repro-readiness is high. Topic counts — Multi-Agent Systems and Negotiation (7), Topic Modeling (5), Logic, Reasoning, and Knowledge (5), Multimodal Machine Learning Applications (4), Natural Language Processing Techniques (3) — show a portfolio bridging agent systems, language technology, and the statistics of evaluation, supported by a compact network of 8 distinct coauthors.
Collaboration pattern
Coauthorship is compact: 8 distinct coauthors across the OpenAlex record, with Yezhou Cheng, Runjia Du, Zeming Liu, Qibai Chen, Hang Lyu, Bojun Lin, and Yilan Wei appearing most frequently (Yilan Wei listed twice in the raw frequent-coauthor data). Both Evolve-catalog papers are 2026 arXiv preprints with middle author positions, suggesting contributions within small teams rather than solo or large-consortium work. The collaboration score of 44 reflects this modest but recurring partner set.
Research keywords
Highlights
Scoped retrieval against cross-family interference in agent memory
Sharp replication limits for hard-budget repeated evaluation
Timeline
- 2014First indexed work appears in the OpenAlex record (1 work).
- 20193 works indexed, the busiest year of the early period.
- 20214 works indexed, the second-highest annual total in the record.
- 2023Peak output year: 7 works indexed.
- 2026Two arXiv preprints enter the Evolve catalog: one on scoped retrieval for agent memory, one on sharp limits for hard-budget repeated evaluation.
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.
- Identity is name-resolved from OpenAlex bibliographic data (ORCID 0009-0007-6817-9747); 'Yankai Zeng' is a name that may be shared by multiple researchers.
- Citation-impact evidence is limited to h-index 5 and i10-index 3; no per-paper citation counts or field-normalized metrics are available.
- Radar scores are unavailable for output, impact, focus, and reproducibility due to insufficient underlying data.
- Only the two 2026 arXiv preprints carry Evolve review verdicts; the remaining ~22 indexed works are bibliometric entries without full-text review.