Michael O'Boyle

University of Edinburgh · United Kingdom

Unclaimed · AI-structured profileOpenAlex

Edinburgh researcher spanning LLM-driven analog circuit design automation and learning-based binary decompilation

papers here

reviewed by Evolve

first-author

corresponding

—

works · OpenAlex

h-index · OpenAlex

Bibliometric figures come from OpenAlex (2026-10-02) 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—

Unscored — the OpenAlex citation record is unpopulated (h-index 0, i10-index 0) and aggregate work counts are absent, so cumulative output cannot be assessed. This axis would measure publication volume and venue tier across the full career.

Impact—

Unscored — citation counts and citation-based indicators are absent from the record (h-index 0, i10-index 0), so impact cannot be estimated. This axis would measure citations, downstream reuse, and the work's influence on its fields.

Focus0/100

Focus scores 0, and the raw data explains why: the field distribution splits exactly 50/50 between LLM-based analog/mixed-signal EDA (circuit design automation) and systems security/programming languages (binary decompilation). The two 2026 catalog papers sit one in each field, so this snapshot shows no dominant specialization. The OpenAlex topic list (VLSI, FPGA, embedded systems, malware detection, photolithography) adds further spread, though it may reflect disambiguation noise.

Collaboration50/100

Collaboration scores 50 on the 0-100 scale. The raw record shows 11 distinct coauthors, and the frequent-coauthor list includes Yijia Hao, Pratibha Verma, Dongxu Guo, Themis Prodromakis, and Rodrigo C. O. Rocha, among others. On both catalog papers O'Boyle is a non-corresponding author — middle on AgenticSizing and last on Echo — indicating contribution within larger teams rather than sole or leading corresponding-author work.

Activity40/100

Activity scores 40. The OpenAlex record shows 3 works in the last three years, with countsByYear listing exactly 3 publications in 2026 — consistent with the two catalog preprints plus one additional record. Recent output is thin but current; longer-term productivity cannot be verified because historical yearly counts are absent from the snapshot.

Reproducibility55/100

Reproducibility scores 55, computed as the mean across 2 papers with reproducibility data. Both catalog papers — AgenticSizing and Echo — are rated medium reproducibility readiness. Evolve's review notes temper the picture: AgenticSizing rests on n=5 evidence with some inconsistent numbers, while Echo reports a 2.43x match-rate gain but with a query-budget asymmetry flagged.

Research summary

This profile is name-resolved from OpenAlex bibliographic data for Michael O'Boyle at the University of Edinburgh. The Evolve catalog holds two 2026 arXiv preprints in strikingly different territories: AgenticSizing, where O'Boyle is a middle author on a large language model multi-agent framework for analog circuit sizing, demonstrated on a 55-transistor LDO where classical optimizers reportedly fail; and Echo, where he is last author on a learning-based matching decompiler using compilation as trusted back-translation, reporting 2.43x GPT-5.6's match rate. The underlying field distribution splits exactly 50/50 between LLM-based analog/mixed-signal EDA and systems security/programming languages, yielding a focus score of 0 — an even spread rather than a specialization. Activity is modest but current: 3 works in the last three years, all recorded in 2026, with 11 distinct coauthors and a collaboration score of 50; neither catalog paper lists him as corresponding author. Reproducibility averages 55 across the 2 papers with repro data, both rated medium readiness. Because the citation record shows an h-index and i10-index of 0, output and impact are left unscored rather than estimated.

Collaboration pattern

The record shows a team-based, non-corresponding pattern: middle author on AgenticSizing and last author on Echo, neither as corresponding author. Eleven distinct coauthors appear in the recent window, with recurring names including Yijia Hao, Pratibha Verma, Dongxu Guo, Cristian Sestito, Christos-Savvas Bouganis, Themis Prodromakis, and Rodrigo C. O. Rocha. The coauthor pool itself straddles both fields — circuits and EDA names alongside decompilation and compilers names — consistent with the even 50/50 field split.

Research keywords

analog circuit sizingLLM agentsmulti-agent frameworkscircuit topology decompositiondesign knowledge reuseneural decompilationmatching decompilationtrusted back-translationbinary analysisprogram synthesiscompiler configuration inference

Highlights

AgenticSizing: LLM multi-agent framework for analog circuit sizing
A middle-author contribution (2026, arXiv) proposing a large language model-based multi-agent framework for analog circuit sizing, with claims of success on a 55-transistor LDO where classical optimizers fail. Evolve's verdict calls it promising but flags n=5 evidence and inconsistent numbers; reproducibility readiness is rated medium.
Echo: matching decompilation via trusted back-translation
A last-author contribution (2026, arXiv) on learning-based matching decompilation that uses compilation as trusted feedback, reporting a 2.43x match rate versus GPT-5.6. Evolve rates the work strong but tempers it with a query-budget asymmetry concern; reproducibility readiness is medium.
A 2026 double pivot across unrelated fields
Both reviewed papers appeared in 2026 and sit in different fields — the field distribution is exactly 50/50 between LLM-based analog/mixed-signal EDA and systems security/programming languages — producing a focus score of 0 and a genuinely dual-track output pattern in a single year.

Timeline

2026AgenticSizing preprint appears on arXiv — a middle-author contribution on an LLM-based multi-agent framework for analog circuit sizing.
2026Echo preprint appears on arXiv — a last-author contribution on learning-based matching decompilation using trusted back translation.
2026OpenAlex records 3 works for the year and 11 distinct coauthors in the recent window; both reviewed papers carry non-corresponding author roles.

Papers in the Evolve catalog

Frequent co-authors

Yijia Hao×1 Pratibha Verma×1 Dongxu Guo×1 Cristian Sestito×1 Christos-Savvas Bouganis×1 Themis Prodromakis×1 Jun Bi×1 Xiangxin Fang×1 Aarsh Chaube×1 José Wesley De Souza Magalhães×1 Rodrigo C. O. Rocha×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 derive from OpenAlex only; no ORCID was provided and the citation record shows an h-index and i10-index of 0, so output and impact axes are unscored.
Both reviewed papers are 2026 arXiv preprints; no peer-reviewed venue information is present in the data.
The OpenAlex topic list (VLSI, FPGA, embedded systems, malware detection, photolithography) does not match the reviewed papers and may reflect name-disambiguation noise in the underlying record.
Author-position roles come from catalog metadata only; career history, awards, and funding fall outside the data scope.