Keyu Chen — works across LLM agent long-term memory and deep learning credit risk early warning
2
papers here
2
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—
Output metrics — total publications, venue types, yearly productivity — are unavailable; current data covers only two catalog-reviewed 2026 arXiv preprints. This axis would measure the volume and consistency of Chen's research output.
Impact—
Impact signals — citations, h-index, downstream adoption — are not present in the input data, so no score can be reported. This axis would measure how widely Chen's work is cited and used beyond the two reviewed 2026 preprints.
Focus0/100
Focus scores 0/100, reflecting an even 50/50 split between two distant fields: natural language processing (LLM agents, long-term memory and personalization) and applied machine learning in finance (credit risk management). With only two reviewed papers, one in each area, there is not yet a dominant research line; the score may shift as a coherent direction emerges across future work.
Collaboration53/100
Collaboration scores 53/100, based on 13 distinct coauthors in the data. On both catalog papers Chen is a middle-position author and not corresponding, consistent with contributing within larger teams rather than leading them. Recurring coauthors include Xinghao Chen, Xiaoyu Shen, Ruifeng Xu, Ruizhi Qiao, and Xing Sun, though paper-level attribution is not available.
Activity—
Activity metrics such as recent publication cadence and ongoing projects cannot be computed from the data, which spans a single year (2026) and two papers. This axis would measure ongoing research tempo and sustained engagement over time.
Reproducibility20/100
Reproducibility scores 20/100, the mean across 2 assessed papers, both rated low reproducibility readiness. The memory paper lacks significance tests, a stated evaluation judge, and released code; the credit-risk paper reports 91.2% accuracy and F1 0.873 but names no baselines and offers no reproduction path. Together they define an early, below-average transparency baseline.
Research summary
Keyu Chen is an early-stage machine learning researcher whose documented output to date sits at the intersection of two distinct areas: natural language processing focused on LLM agents with long-term memory and personalization, and applied machine learning in finance for credit risk management. Evolve's catalog currently includes two 2026 arXiv papers with Chen as a middle-position, non-corresponding author. The first, 'Disentangling Long-Term Memory via Latent Neuro-Symbolic Reasoning', proposes query-conditioned latent graphs drawing on sparse autoencoders, latent-space reasoning, and graph neural networks for personalized agents; Evolve's review notes large reported gains on personalization benchmarks but no significance tests, an unstated judge, and no code release. The second, a credit risk early warning system, integrates multi-source heterogeneous data with a GCN+BiLSTM+attention architecture and real-time stream processing, reporting 91.2% accuracy, F1 0.873, and a 28-day lead time; the review flags unnamed baselines, no statistical tests, and irreproducibility. Bibliometric signals reflect this early profile: a reproducibility score of 20 (mean across both reviewed papers, each rated low readiness), a focus score of 0 from an even 50/50 field split, and a collaboration score of 53 with 13 distinct coauthors. Output, impact, and activity metrics are not yet available in the underlying data, and institution, country, and ORCID remain unresolved.
Collaboration pattern
Chen's recorded network includes 13 distinct coauthors, yielding a mid-range collaboration score of 53. Recurring names include Cai Ke, Xinghao Chen, Xiaoyu Shen, Siyu An, Junnan Dong, Ruifeng Xu, Ruizhi Qiao, Xing Sun, LiYang Wang, Zhen Zhong, and Zhen Tian. On both catalog papers Chen sits in middle author positions without corresponding status, suggesting a contributor role within multi-author teams; paper-level co-author mappings are not in the data.
Research keywords
Highlights
Query-conditioned latent graphs for agent long-term memory
GCN+BiLSTM+attention credit risk early warning system
Timeline
- 2026Co-authored (middle position, non-corresponding) 'Disentangling Long-Term Memory via Latent Neuro-Symbolic Reasoning' on arXiv, in LLM agent long-term memory and personalization.
- 2026Co-authored (middle position, non-corresponding) the deep learning credit risk early warning system paper on arXiv, in applied machine learning for finance.
- 2026Both 2026 papers entered the Evolve catalog with low reproducibility readiness ratings; the bibliometric snapshot records 13 distinct coauthors and an even 50/50 split across the two fields.
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.
- Name-resolution note: 'Keyu Chen' is a common name; this profile is resolved from public bibliographic data, and institution, country, and ORCID are not confirmed.
- Bibliometrics come from a limited snapshot covering only two catalog-reviewed 2026 arXiv preprints; the output, impact, and activity axes lack underlying data.
- The supplied frequent-coauthor list contains an entry identical to the researcher's own name, suggesting possible entity-resolution noise in the co-author graph.
- The reproducibility score (mean 20) is based on Evolve's review of only 2 papers; it reflects those reviews, not a full audit of all work.