Vivek Gupta

Arizona State University · United States

Unclaimed · AI-structured profileOpenAlex

Arizona State University NLP researcher: clinical QA benchmarks, privacy-hardened evidence, federated LLM agents

papers here

reviewed by Evolve

first-author

corresponding

—

works · OpenAlex

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: this OpenAlex snapshot records h-index 0 and i10-index 0, with only 23 works counted, all in 2026. This axis would measure publication volume and cumulative productivity across the full career once coverage is complete.

Impact—

Citation impact cannot be scored yet: h-index and i10-index are recorded as 0 in this OpenAlex identity, indicating insufficient citation coverage rather than a measured result. This axis would measure citation-based influence such as h-index, citations per paper, and citedness.

Focus0/100

Scores 0/100 on focus, consistent with a deliberately broad portfolio rather than absence of work: the raw field distribution splits evenly (0.33 each) across clinical NLP, NLP with LLM agents and federated learning, and general NLP, while OpenAlex topics span topic modeling (17), search behavior (10), multimodal ML (7), adversarial robustness (5), and NLP techniques (4).

Collaboration54/100

Scores 54/100 for collaboration: 14 distinct co-authors appear in the recent three-year window, and a 12-person frequent-collaborator list spans the clinical NLP strand and the federated-agents strand. On all three cataloged 2026 papers Gupta holds the last author slot, a pattern consistent with a coordinating role on joint system and benchmark projects.

Activity93/100

Scores 93/100: 23 works fall in the recent three-year window, and the snapshot's countsByYear places all 23 in 2026 alone. The three cataloged papers (SentryLine, CIPHER, SYNAPSE) all carry 2026 arXiv dates, indicating a dense current burst of benchmark and system building.

Reproducibility70/100

Scores 70/100, the mean readiness across 3 reviewed papers. Evolve rates SYNAPSE as high readiness — 'well-controlled, unusually honest' — while SentryLine and CIPHER are medium, both relying on LLM-as-judge evaluation and, per the verdicts, lacking significance tests, which tempers but does not erase reproducibility.

Research summary

Vivek Gupta is a researcher at Arizona State University whose recent work sits at the intersection of clinical natural language processing and the systems side of NLP — retrieval-augmented question answering, privacy-preserving evaluation, and federated infrastructure for LLM agents. His profile is name-resolved from OpenAlex bibliographic data: the snapshot records 23 works in the recent three-year window (activity 93/100), 14 distinct co-authors, and an even three-way split (0.33 each) across clinical NLP, NLP with LLM agents and federated learning, and general NLP. Topic Modeling (17) and Information Retrieval and Search Behavior (10) lead his topic profile, with further work in multimodal machine learning (7), adversarial robustness (5), and NLP techniques (4). Three 2026 arXiv papers in the Evolve catalog, each with Gupta as last author, illustrate the agenda. SentryLine builds evidence-grounded question answering over evolving oncology-care documents; Evolve calls it a strong system-plus-benchmark but notes missing significance tests and judge-family bias. CIPHER benchmarks cross-record inference over privacy-hardened records with 3,427 questions and SQL provenance, though findings hinge on a lenient judge. SYNAPSE introduces typed federated artifacts for tool routing across frozen, heterogeneous agents — rated 'well-controlled, unusually honest,' with high reproducibility readiness. Mean reproducibility across the three reviewed papers is 70/100.

Collaboration pattern

Collaboration scores 54/100 with 14 distinct recent co-authors and a 12-person frequent-collaborator list spanning both strands: NLP-side collaborators such as Tampu Ravi Kumar, Gaurav Najpande, Muhammad Ali Khan and Kaneez Zahra Rubab Khakwani, alongside clinical-affiliated names including M Bassam Sonbol, Irbaz Bin Riaz and Abhijit Chakraborty. Across all three cataloged 2026 papers Gupta takes the last author slot, consistent with a coordinating role on joint system and benchmark projects.

Research keywords

retrieval-augmented generationclinical question answeringoncology informaticsliving clinical guidelinesLLM-as-judge evaluationbenchmark constructionhybrid table-text QAprivacy-preserving de-identificationtext-to-SQL provenancefederated learningLLM agents and tool routingtemporal drift

Highlights

SentryLine: evidence-grounded QA over evolving oncology documents
A 2026 arXiv system-plus-benchmark for guideline-drift question answering in oncology care, with Gupta as last author. Evolve's verdict credits a strong construction for living guidelines but flags missing significance tests, judge-family bias, and unconditioned baselines tempering headline gains; reproducibility readiness is medium.
CIPHER: cross-record inference over privacy-hardened records
A 2026 arXiv benchmark for question answering over de-identified hybrid table-text evidence, built with 3,427 questions and SQL provenance. Evolve finds the findings credible but dependent on a lenient judge and a one-model restoration ablation; reproducibility readiness is medium.
SYNAPSE: typed federated artifacts for tool routing
A 2026 arXiv paper proposing typed federated artifacts that share tool-routing knowledge across frozen, heterogeneous LLM agents, with conflict resolution and differential privacy. Evolve calls it well-controlled and unusually honest, though gains trace to typed rendering rather than federation itself; readiness is high.

Timeline

2026Last-author arXiv paper SentryLine: evidence-grounded question answering over evolving documents in oncology care.
2026Last-author arXiv benchmark CIPHER: cross-record inference over privacy-hardened evidence records.
2026Last-author arXiv paper SYNAPSE: typed federated artifacts sharing tool-routing knowledge across frozen, heterogeneous LLM agents.

Papers in the Evolve catalog

Frequent co-authors

Tampu Ravi Kumar×1 Gaurav Najpande×1 Muhammad Ali Khan×1 Kaneez Zahra Rubab Khakwani×1 Karan Kathuria×1 Shorya Azriel Moses×1 Yuvraj Kalia×1 M Bassam Sonbol×1 Irbaz Bin Riaz×1 Abhijit Chakraborty×1 Ni Trieu×1 Suparno Roy Chowdhury×1

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 an OpenAlex identity snapshot only; output and impact are left unscored because h-index and i10-index are recorded as 0.
The snapshot's countsByYear contains only 2026 (23 works), so multi-year productivity trends cannot be assessed.
No ORCID is linked and 'Vivek Gupta' is a common name, so this profile is name-resolved from public bibliographic data and may merge or split same-name records.
Reproducibility is the mean readiness over the 3 Evolve-reviewed catalog papers, not a full-corpus measure.