Purdue University researcher spanning traffic management and safety, LLM agent memory, and evaluation methodology
2
papers here
0
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
12
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—
Not yet scored. The record provides an h-index of 12 and an i10-index of 16, but venue-normalized publication data is insufficient. This axis would measure total research output and the quality of publication venues across the whole career.
Impact—
Not yet scored. Per-work citation counts are absent from the provided snapshot, so influence cannot be graded. This axis would measure citation impact, including the h-index trajectory and the reach of the most-cited works.
Focus—
Not yet scored. Topic labels exist in the record, but coverage is too thin to grade concentration. This axis would measure how tightly the publication record clusters around a coherent research agenda over time.
Collaboration44/100
Scores 44 on the collaboration axis. The OpenAlex record lists 8 distinct coauthors, with recurring partners such as Yezhou Cheng, Zeming Liu, Qibai Chen, Hang Lyu, Yankai Zeng, Bojun Lin, and Yilan Wei. Both 2026 arXiv papers in the Evolve catalog list a middle author position without corresponding status, consistent with contributing inside team-led projects rather than solo or senior-led work.
Activity40/100
Scores 40 on the activity axis. The counts-by-year record spans 2013 to 2025 and peaks at seven works in each of 2020 and 2021, then tapers to five in 2022, three in 2023, one in 2024, and two in 2025; only 3 works fall in the most recent three-year window. The two 2026 arXiv preprints suggest output has not fully stalled.
Reproducibility—
Not yet scored at portfolio level. Evolve's two catalog reviews rate repro-readiness medium and high, but the sample is too small for an axis score. This axis would measure code, data, and replication evidence across all works.
Research summary
Runjia Du is a researcher at Purdue University (West Lafayette, US) whose public bibliographic record, resolved via OpenAlex, spans transportation and machine-learning evaluation. The recorded topic profile is dominated by traffic research: traffic control and management (17 topic-assigned works), traffic prediction and management techniques (11), autonomous vehicle technology and safety (10), transportation planning and optimization (6), and traffic and road safety (5). Aggregate indicators are modest: an h-index of 12 and an i10-index of 16, with counts-by-year spanning 2013 to 2025 and peaking at seven works in each of 2020 and 2021 before tapering to one in 2024 and two in 2025; only 3 works fall in the most recent three-year window. The Evolve catalog adds two 2026 arXiv preprints of a different flavor: one on scoped retrieval preventing cross-family interference in LLM agent memory, and one proving sharp replication limits for hard-budget repeated LLM evaluation. Both list the researcher in a middle author position. Evolve's verdicts credit solid paired evidence in the first and matching designs with equal-budget evidence in the second, while flagging single-benchmark and regime-dependent caveats. The split between the transportation-heavy topic record and the recent LLM-methods preprints is noted but unverified.
Collaboration pattern
The record lists 8 distinct coauthors, with recurring collaborators including Yezhou Cheng, Zeming Liu, Qibai Chen, Hang Lyu, Yankai Zeng, Bojun Lin, and Yilan Wei. The collaboration score of 44 reflects a modest, partner-driven network. Both 2026 arXiv papers in the Evolve catalog place this researcher in a middle author position without corresponding status, indicating contribution within team-led projects rather than solo or senior-lead authorship.
Research keywords
Highlights
Scoped retrieval against cross-family interference in agent memory
Sharp limits for honest uncertainty in hard-budget LLM evaluation
Timeline
- 2013Earliest recorded publication year in the OpenAlex counts-by-year data, with one work.
- 2020Output reaches the joint peak of the record: seven works, matched only by 2021.
- 2021Seven works recorded again, extending the 2020–2021 stretch as the most productive span in the record.
- 2022Five works recorded as the yearly count begins to taper (three in 2023, one in 2024).
- 2026Two arXiv preprints enter the Evolve catalog — scoped retrieval for agent memory and sharp limits for hard-budget repeated evaluation — both in a middle author position.
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 are drawn from OpenAlex only (identity: openalex); citation and topic coverage may be incomplete.
- This profile is name-resolved from public bibliographic data; the split between transportation-heavy topics and recent LLM-methods preprints may reflect a research shift or an identity-disambiguation issue, which cannot be verified here.
- Only two papers carry Evolve catalog reviews; portfolio-level output, impact, focus, and reproducibility axes remain unscored for lack of data.
- Both cataloged papers are 2026 arXiv preprints; peer-reviewed venue status is not confirmed in the provided data.