Yuran Chen
Dongbei University of Finance and Economics · China
NLP researcher blending LLM-as-a-judge calibration and evidence fusion to measure value and moral judgment at scale
2
papers here
2
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
0
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-10-16) 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 available yet: the record's works data is too thin to compute one. This axis would measure publication volume and its distribution across venues and years; only 2026 arXiv entries are visible so far.
Impact—
No impact score is available yet. OpenAlex reports h-index 0 and i10-index 0, which reflects a brand-new record rather than measured influence; this axis would track citations and downstream uptake once they accrue.
Focus0/100
Focus scores 0/100, and the raw data supports the reading of a broad portfolio: the field split is exactly 50/50 between LLM alignment and evaluation and NLP/computational social science, and topic counts stretch from topic modeling (4) and multimodal machine learning (3) into robotics (reinforcement learning 1, manipulation 1). This is dispersed, early-stage exploration rather than a narrow specialty.
Collaboration62/100
Collaboration scores 62/100, drawn from 21 distinct coauthors on the OpenAlex record. The frequent-partner list includes Yuhe Wu, Guangyu Wang, Yujie Chen, Jiatong Zhang, Yutong Zhang, Xiyin Cheng, Wenpeng Cao, Zhuang Liu, Guang Zhang, Rui Qian, and Yuanchao Zhu, and both reviewed papers sit in middle authorship positions — consistent with small, multi-author arXiv teams rather than solo work.
Activity57/100
Activity scores 57/100. The countsByYear series contains a single entry — 6 works in 2026 — and recent3yWorks is likewise 6, so all indexed output lands in one burst year with no earlier record. The moderate score reflects a short observable window rather than a slowdown; this axis can only be judged properly once more years accumulate.
Reproducibility55/100
Reproducibility scores 55/100, the mean readiness rating across the 2 papers Evolve has reviewed (reproPapers = 2). Both carry a 'medium' tag: the TEF fusion method is simple and effective but leans on synthetic data and weak baselines, while the Narrative Captivity benchmark is substantial (5,078 scenarios) yet has a contested DPO attribution step.
Research summary
Yuran Chen is a China-based NLP researcher affiliated with Dongbei University of Finance and Economics whose 2026 output concentrates on two linked questions: how to measure long-text value orientation reliably, and how multi-turn conversation shapes what large language models judge. In the computational social science strand, the reviewed paper on Tempered Evidence Fusion (TEF) entropy-weights sentence-level log-odds from multiple LLM judgments and reports roughly 4.5 accuracy points over naive voting, though weak baselines and synthetic data temper confidence. In the alignment-and-evaluation strand, the Narrative Captivity benchmark spans 5,078 multi-turn scenarios and finds one-sided narration shifts LLM moral judgment by about 25 percentage points, with the DPO-based attribution flagged as strained. Bibliometrically, the OpenAlex record is nascent: h-index and i10-index are both 0, all 6 indexed works fall in 2026, and 21 distinct coauthors appear (collaboration 62/100; activity 57/100). Topic counts span topic modeling (4), NLP techniques (3), multimodal machine learning (3), and robotics entries, explaining a focus score of 0/100. Both reviewed papers list Chen in a middle authorship position with medium reproducibility ratings (mean 55/100). This is an early-stage profile built from OpenAlex bibliometrics plus two Evolve-reviewed preprints.
Collaboration pattern
Yuran Chen works in small multi-author teams: the OpenAlex record lists 21 distinct coauthors, and the Evolve frequent-coauthor set includes Yuhe Wu, Guangyu Wang, Yujie Chen, Jiatong Zhang, Yutong Zhang, Xiyin Cheng, Wenpeng Cao, Zhuang Liu, Guang Zhang, Rui Qian, and Yuanchao Zhu. On both reviewed arXiv preprints Chen holds a middle authorship position without corresponding-author status, suggesting a contributing rather than leading role at this stage.
Research keywords
Highlights
Entropy-weighted evidence fusion beats sentence-level voting
5,078-scenario benchmark shows narration sways LLM moral judgment
Timeline
- 2026Reviewed preprint on Tempered Evidence Fusion (TEF) for long-text value measurement appears on arXiv, with Chen in a middle authorship position.
- 2026Reviewed preprint 'Caught in the Story' introduces the Narrative Captivity benchmark of 5,078 multi-turn scenarios for LLM conversations.
- 2026OpenAlex countsByYear records 6 works in 2026 — the only year present in the snapshot, with h-index and i10-index still at 0.
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 come from OpenAlex only (identity: openalex); no ORCID is linked, and 'Yuran Chen' is a common name, so this profile is name-resolved from public bibliographic data.
- Output and impact dimensions are unscored because the record is nascent — h-index and i10-index are both 0, and all 6 indexed works fall in 2026.
- Qualitative verdicts rest on just 2 Evolve-reviewed arXiv preprints, both with middle authorship and neither corresponding-authored.
- The focus score of 0 is computed from a thin field/topic distribution and should be read as 'broad, early-stage portfolio', not a definitive specialization measure.