NLP researcher at Hong Kong PolyU: citation faithfulness verification and evidence-grounded clinical mental-health NLP.
2
papers here
2
reviewed by Evolve
1
first-author
0
corresponding
—
works · OpenAlex
—
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-10-13) 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 is not scored: no publication totals or venue record beyond two Evolve-reviewed arXiv papers is available. This axis would measure annual paper volume and venue placement; the current two-paper sample cannot support an estimate.
Impact—
Impact is not scored: no citation counts or h-index are provided for this profile. This axis would measure citations, uptake, and influence of the work; with 2026 arXiv papers too recent to have accumulated citations, the axis stays empty.
Focus0/100
Focus scores 0 because Yanling Li's reviewed output is split exactly evenly: the field distribution is 50% natural language processing and 50% clinical natural language processing (computational mental health). Rather than one narrow specialty, the profile spans general NLP methodology — ensembles and calibration — and clinical applications such as suicide-risk detection. The even split signals breadth; a larger reviewed corpus would show whether one direction consolidates.
Collaboration28/100
Collaboration scores 28, with only 3 distinct coauthors recorded across the reviewed papers — Zirui Li, Kaolanglang Gao, and Mingyu Wan. This is a compact footprint: small teams, recurring partners, and no large multi-institution consortium visible in the data. Yanling Li is first author on one reviewed paper and a middle author on the other, with no corresponding-author role recorded, consistent with an early-stage contributor working within a close-knit group.
Activity—
Activity is not scored: no year-by-year publication stream or ORCID-verified history is available. This axis would measure publishing cadence over time; it requires a fuller bibliographic record than the two 2026 papers and absent identifiers used here.
Reproducibility55/100
Reproducibility scores 55, a mean across the 2 reviewed papers, each rated medium reproducibility readiness. The NLPCC 2026 shared-task paper supports its 82.99 runner-up score with thorough ablations and honest limitation statements, though key conclusions rest on out-of-fold diagnostics without significance testing; the suicide-risk detection paper reports small gains validated by confirmation checks. These are encouraging methodological habits, but medium readiness means replication still depends on unreleased details.
Research summary
Yanling Li is a natural language processing researcher at Hong Kong Polytechnic University whose reviewed work sits at the intersection of general NLP methodology and clinical natural language processing for computational mental health — the Evolve field distribution assigns each of these two areas exactly half of the reviewed output. Two 2026 arXiv papers make up the current catalog. As first author, Li led the NLPCC 2026 Task 10 entry on citation-level faithfulness verification, where DeBERTa ensembles with class-wise calibration earned a Track 2 runner-up finish with an 82.99 overall score, backed by thorough ablations and explicit limitation statements. As a middle author, Li contributed to a suicide-risk detection study proposing heterogeneous readouts (HRD) that replace final-token classification margins with latent readouts of LLM internal representations, achieving small but confirmation-validated gains, with probing classifiers and QLoRA fine-tuning among the underlying techniques. A recurring theme across both papers is evidence: retrieving and calibrating evidence for citation faithfulness, and extracting evidence to ground risk decisions. The bibliometric footprint is compact — three distinct coauthors (Zirui Li, Kaolanglang Gao, Mingyu Wan) — and Evolve's reproducibility rating averages 55 across the two papers, both at medium readiness. Output, impact, and activity dimensions are not yet scored because citation counts, publication totals, and a verified publication history are unavailable; this is an early-stage profile built from a small, well-documented sample.
Collaboration pattern
All reviewed work comes from small teams: only 3 distinct coauthors — Zirui Li, Kaolanglang Gao, and Mingyu Wan — appear across the two papers, giving a low collaboration score of 28. Yanling Li holds a first-author role on the NLPCC shared-task paper and a middle-author role on the suicide-risk detection study, with no corresponding-author role recorded. The pattern suggests a tight, recurring trio rather than a broad multi-institution network; the coauthor pool may grow as output expands.
Research keywords
Highlights
NLPCC 2026 Track 2 runner-up on citation faithfulness verification
Heterogeneous readouts move beyond final-token classification for suicide risk detection
An evidence-centered methodological thread across two fields
Both reviewed papers center on evidence — retrieval and calibration for citation faithfulness, and evidence extraction for grounded risk decisions — which explains the even 50/50 split between natural language processing and clinical NLP (computational mental health) in the field distribution.
Timeline
- 2026First-authored the NLPCC 2026 Task 10 entry on citation-level faithfulness verification, combining DeBERTa ensembles with class-wise calibration.
- 2026Co-authored (middle position) the heterogeneous-readouts study on evidence-grounded suicide risk detection.
- 2026Both papers entered the Evolve catalog, spanning natural language processing and clinical NLP (computational mental health), with medium reproducibility readiness and a reproducibility mean of 55.
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.
- Yanling Li is a relatively common name; this profile is resolved from public bibliographic data (two Evolve-reviewed arXiv papers) without ORCID verification, so identity beyond the listed affiliation is not guaranteed.
- Bibliometric aggregates cover only the two reviewed papers; no citation counts, h-index, or publication totals were available, so output, impact, and activity remain unscored.
- Both catalog papers are 2026 arXiv entries; venue and peer-review status beyond the NLPCC 2026 shared-task result is not recorded.
- The reproducibility score (mean 55) is based on only 2 papers rated medium; a broader corpus could shift this rating substantially.