University of Toronto researcher linking grounded legal reasoning datasets to multi-agent LLM bioinformatics systems.
2
papers here
2
reviewed by Evolve
1
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—
Output is not scored yet: the OpenAlex record shows h-index 0 and i10-index 0, with works appearing only in 2026, so there is no stable volume signal. This axis would measure publication counts, venue spread, and authorship roles over time.
Impact—
Impact is not scored yet: with i10-index 0 and a record that begins in 2026, no citation signal exists to evaluate. This axis would measure citations, downstream adoption, and uptake of the two 2026 preprints as they accumulate.
Focus0/100
Focus scores 0, and the raw field distribution explains why: work splits exactly 50/50 between legal NLP (computational law) and AI agents for bioinformatics, while OpenAlex topics scatter across five areas — topic modeling (2 works), NLP techniques, social robot HRI, AI planning, and adversarial robustness (1 each). This is a broad or not-yet-settled research portfolio with no dominant specialization visible in the data.
Collaboration68/100
The collaboration axis scores 68, supported by 29 distinct coauthors in the recent window and a frequent-coauthor list of 12 names including Wantong Huo, Jonathan H. Chan, and Zhouyang Yuan. Both catalog papers are multi-author arXiv reports — GRACE with a first-author position, Bioinfoysis in a middle position — with no corresponding-author record, suggesting active participation in larger team projects rather than solo research.
Activity47/100
Activity scores 47. OpenAlex records 4 works in the recent three-year window, and countsByYear places all of them in 2026, so measurable output is concentrated in a single year — too short a record to establish a sustained publishing cadence. Both catalog papers appeared on arXiv in 2026. The moderate score reflects genuinely recent output volume rather than a long publication track record.
Reproducibility38/100
Reproducibility scores 38, computed as the mean readiness (reproMean 38) across 2 Evolve-reviewed papers. GRACE is rated medium, thanks to a grounded 1,915-example Canadian legal dataset and a solid verification pipeline; Bioinfoysis is rated low, with abstract-body numeric conflicts and audit-informed skills flagged for caution. The below-midpoint score means reproducing this work currently requires care and likely direct author clarification.
Research summary
Jiakang Xu is an early-career researcher at the University of Toronto whose OpenAlex profile (ORCID 0009-0007-9776-3694) is name-resolved from public bibliographic data and currently compact: 4 works recorded in 2026, h-index 0, i10-index 0, and 29 distinct coauthors. The field distribution splits exactly 50/50 between legal NLP (computational law) and AI agents for bioinformatics, making this a genuinely two-domain record rather than a specialized one. Two cataloged 2026 arXiv papers anchor the evidence. GRACE: Grounded Adversarial Reasoning over Canadian Law, with Xu as first author, contributes a grounded 1,915-example Canadian legal dataset and a citation-verification pipeline; Evolve's verdict credits the dataset and pipeline but cautions that evaluation against its own teacher limits correctness claims. Bioinfoysis Technical Report, a middle-author position, reports 82.4% state of the art on BixBench with strong ablations, though numeric conflicts between abstract and body temper the claim. Radar scores capture the shape: collaboration 68, activity 47, reproducibility 38 (mean readiness across the two reviewed papers), and focus 0, reflecting an even split across five scattered OpenAlex topics. Output and impact are unscored because citation indicators remain at zero this early. This is an early-stage, team-oriented profile whose direction — grounded legal reasoning datasets and benchmark-driven LLM agents — is visible but not yet consolidated.
Collaboration pattern
Collaboration scores 68 on the strength of 29 distinct coauthors in the recent window and 12 recurring partners, including Wantong Huo, Jonathan H. Chan, Zhouyang Yuan, and Xiaoqing Yan. Both cataloged papers are multi-author arXiv works — GRACE with a first-author position, Bioinfoysis in a middle position — and neither lists Xu as corresponding author. The pattern suggests active contributions within larger team projects rather than small-group or solo leadership so far.
Research keywords
Highlights
Grounded Canadian legal dataset with verification pipeline
First-authored entry into computational law
BixBench state-of-the-art claim from a multi-agent harness
Timeline
- 2026OpenAlex records 4 works in 2026 — the entire visible output of this profile to date — with h-index and i10-index still at 0.
- 2026First-author arXiv paper GRACE: Grounded Adversarial Reasoning over Canadian Law introduces a 1,915-example grounded Canadian legal dataset with a citation-verification pipeline.
- 2026Co-authors the Bioinfoysis Technical Report (middle position), a multi-agent bioinformatics harness claiming 82.4% on BixBench with strong ablations.
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.
- Bibliometric indicators come from OpenAlex only; the profile is name-resolved from public bibliographic data and should not be treated as a verified biography.
- Output and impact are unscored (dimMissing): h-index 0 and i10-index 0 with works appearing only in 2026 leave no citation signal to evaluate.
- Only the two 2026 arXiv preprints carry Evolve reviews; the reproducibility score (mean 38) reflects exactly those two papers, not the full output.
- All timeline evidence comes from a single year (2026); no earlier publication history is present in the input data.