Anadolu University researcher on fractal zero-storage neural parameterization and edge-AI decision systems
2
papers here
2
reviewed by Evolve
2
first-author
0
corresponding
—
works · OpenAlex
5
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-10-08) 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—
The output axis measures publication volume and productivity over time, but this snapshot lacks the complete work-count data needed to score it; only fragmentary per-year counts (21 works in 2026, 4 in 2012) are recorded, so no output score is assigned.
Impact—
The impact axis measures citation influence such as per-paper citations and field-weighted uptake; apart from a recorded h-index of 5 and i10-index of 2, no citation distribution is available in this snapshot, so no score is assigned.
Focus0/100
The focus score of 0 reflects unusually wide topical dispersion: OpenAlex field weights split evenly at 0.5 for edge AI / efficient inference systems and 0.5 for machine learning (neural network parameterization) / nonlinear dynamics, and the top five topic clusters — Advanced Memory and Neural Computing (10), Neural Networks and Applications (9), Face and Expression Recognition (7), Face recognition and analysis (6), Complex Systems and Dynamics (5) — cut across several communities rather than one concentrated line.
Collaboration36/100
Collaboration scores 36 on thin co-authorship: only 5 distinct co-authors appear in the recent window, and the most frequent named collaborators are Zerrin Dağlı and Dağhan Dağlı, whose entries recur with a shared surname. Both Evolve-cataloged 2026 arXiv papers list the researcher as first author without corresponding-author designation. The co-author list also repeats the researcher's own name, a pattern suggesting name-disambiguation noise in the underlying OpenAlex record.
Activity91/100
Activity scores 91, the strongest axis: the record lists 22 works in the last three years, and per-year counts show 21 works in 2026 alone — versus 1 in 2024, 1 in 2021, 4 in 2012 and 2 in 2011. Both 2026 arXiv preprints in the Evolve catalog are first-authored, consistent with this burst of independent output; the concentration in a single year is the defining feature of the profile.
Reproducibility55/100
Reproducibility scores 55, a mean across 2 assessed papers, both rated medium readiness. Evolve's reviews flag concrete gaps: the zero-storage synthesis paper supports its Mandelbrot-seed weight claim with a single Two-Moons toy test against a 16-byte baseline and leaves headline mechanisms untested, while the decision-map paper's 93.1% headline figure risks in-sample seed tuning and its 0% injection bypass rests on only 10 attack classes.
Research summary
Volkan Dağlı is a researcher at Anadolu University in Türkiye whose record, name-resolved from OpenAlex public bibliographic data, shows an unusually concentrated output pattern: 22 works in the last three years, 21 of them logged in 2026, against earlier annual counts of 2 (2011), 4 (2012), 1 (2021) and 1 (2024). Current work sits in two fields weighted equally at 0.5 each — edge AI / efficient inference systems and machine learning centred on neural-network parameterization and nonlinear dynamics. Two 2026 arXiv preprints, both first-authored, propose generating neural-network weights procedurally from Mandelbrot-set seeds (zero-storage synthesis) and a weight-free fractal decision map for real-time edge triage with claimed prompt-injection resistance. Evolve's reviews find both promising but under-evidenced: support rests on a single Two-Moons toy test against a 16-byte baseline, while headline figures of 93.1% accuracy and 0% injection bypass are exposed to in-sample seed tuning and a narrow 10-attack-class benchmark. Earlier topic clusters — advanced memory and neural computing, face and expression recognition, complex systems — indicate a broader prior agenda. Citation indicators are thin (h-index 5, i10-index 2), co-authorship is narrow (5 distinct co-authors), and the record carries visible disambiguation noise.
Collaboration pattern
Collaboration is narrow: the record lists only 5 distinct co-authors, and the most frequent collaborator names are Zerrin Dağlı and Dağhan Dağlı, each recurring across entries. Both Evolve-cataloged 2026 arXiv papers are first-authored without corresponding-author designation, pointing to largely independent work. The co-author list also repeats the researcher's own name, so collaborator statistics should be read as provisional until the OpenAlex identity graph is cleaned.
Research keywords
Highlights
Zero-storage neural synthesis from Mandelbrot seeds
Weight-free fractal decision map for real-time edge triage
Timeline
- 2011Earliest works appear in the OpenAlex record, with 2 publications logged.
- 20124 publications logged — the highest recorded annual count before 2026.
- 20211 publication logged in a low-output stretch.
- 20241 publication logged, immediately before the 2026 output surge.
- 202621 publications logged — the largest annual count in the record; two first-authored arXiv preprints on Mandelbrot-based zero-storage neural synthesis and fractal edge-triage decision maps enter the Evolve catalog.
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.
- All bibliometrics are resolved from OpenAlex only; no Scopus, Web of Science, or publisher data is included.
- Annual work counts are sparse and gapped (2013–2020, 2022–2023 and 2025 have no recorded works), so long-run productivity trends cannot be established.
- The frequent-coauthor list includes the researcher's own name and repeated same-surname entries, indicating likely name-disambiguation noise; treat identity-level statistics as provisional.
- Citation-impact data is limited to h-index 5 and i10-index 2; per-paper citation distributions were unavailable, so the output and impact axes remain unscored.