Online learning and algorithmic game theory researcher at Google, rising output in minimax-regret theory
2
papers here
2
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
3
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-10-15) 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 is not scored: this snapshot lacks full publication counts, venue breakdown, and output volume beyond the two reviewed papers. It would measure the scale, consistency, and productivity of the published record across years and venues.
Impact—
Impact data is insufficient to score this axis; the only citation figures provided are h-index 3 and i10 index 1. It would measure citation influence, uptake, and recognition of the researcher's work across the community.
Focus0/100
Focus scores 0, signalling breadth rather than a single niche. The Evolve-reviewed work splits evenly (0.5/0.5) between online learning / learning theory and online learning and optimization theory, while the OpenAlex topic profile spans five clusters: Machine Learning and Algorithms (23 works), Auction Theory and Applications (17), Game Theory and Voting Systems (12), Optimization and Search Problems (11), and Game Theory and Applications (10). Thematically these still form one theory-driven agenda.
Collaboration50/100
Collaboration scores 50, with 11 distinct coauthors in the snapshot. Recurring partners include Vineet Gupta, Yanchen Jiang, Christopher Liaw, Aranyak Mehta, Yang Cai, Di Wang, Anupam Gupta, Guru Guruganesh, and Renato Paes Leme — a cluster consistent with industry theory research. Both reviewed papers are middle-author, non-corresponding contributions, fitting work within larger theory teams rather than solo or mass-collaboration projects.
Activity100/100
Activity scores 100, the strongest axis in this profile. OpenAlex counts by year climb from 1 (2019), 1 (2020), and 2 (2021) through 6 (2022) and 4 (2023) to 10 (2024), 7 (2025), and a peak of 16 (2026), with 33 works in the most recent three-year window — a steep, sustained ramp in theoretical output.
Reproducibility100/100
Reproducibility scores 100 (mean 100 across the 2 reviewed papers). Evolve's human verification found the anytime expert-advice proof correct, matching the fixed-horizon regret constant though with asymptotic overhead that vanishes very slowly, and the trace-power-potential proof for O(d)-regret online inverse optimization short, self-contained, and correct; both papers carry high reproducibility readiness.
Research summary
Grigoris Velegkas is a researcher at Google (United States) working at the intersection of online learning, optimization theory, and algorithmic game theory; this profile is name-resolved from public bibliographic data (OpenAlex identity, ORCID 0000-0001-7148-0548). Activity is the standout axis: OpenAlex counts by year rise from 1 (2019) to 2 (2021), 6 (2022), 10 (2024), and a peak of 16 (2026), with 33 works in the most recent three-year window. The topic profile spans Machine Learning and Algorithms (23 works), Auction Theory and Applications (17), Game Theory and Voting Systems (12), Optimization and Search Problems (11), and Game Theory and Applications (10), while the Evolve-reviewed papers split evenly (0.5/0.5) between online learning / learning theory and online learning and optimization theory. Two 2026 arXiv papers have been human-verified by Evolve: an anytime expert-advice algorithm matching the fixed-time regret constant, and a deterministic O(d)-regret bound for proper online inverse optimization; both score 100 on reproducibility readiness (mean 100 across 2 papers). Citation-based impact remains thin — h-index 3, i10 index 1 — so output and impact axes lack sufficient data. Collaboration is moderate (score 50, 11 distinct coauthors), with recurring partners including Vineet Gupta, Yanchen Jiang, Christopher Liaw, Aranyak Mehta, Yang Cai, Di Wang, Anupam Gupta, and Renato Paes Leme.
Collaboration pattern
Velegkas publishes within mid-sized theory teams: 11 distinct coauthors appear in the snapshot, and both reviewed papers are middle-author, non-corresponding contributions. Recurring partners include Vineet Gupta, Yanchen Jiang, Christopher Liaw, Aranyak Mehta, Yang Cai, Di Wang, Anupam Gupta, Guru Guruganesh, and Renato Paes Leme, a cluster consistent with Google's algorithmic game theory and online learning research. The collaboration score of 50 reflects a moderate, concentrated network rather than solitary or mass-collaboration work.
Research keywords
Highlights
Anytime expert-advice regret matches the fixed-time constant
Deterministic O(d) regret for online inverse optimization
Human-verified theory with top reproducibility readiness
Both reviewed papers score 100 on reproducibility (mean 100 across 2 papers), with Evolve verdicts confirming correct, self-contained proofs — a hallmark of this researcher's parameter-free, minimax-oriented theory style.
Timeline
- 2019First works appear in the OpenAlex record: 1 indexed work in 2019, followed by 1 in 2020 and 2 in 2021 — the early build-up phase.
- 2022Indexed output more than doubles to 6 works in 2022, the first clear step up in the record.
- 202410 works indexed in 2024, then the highest annual count in the OpenAlex record.
- 2026Peak activity: 16 works indexed in 2026, including both Evolve-reviewed arXiv papers — the anytime expert-advice regret result and the O(d)-regret online inverse optimization paper.
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.
- Bibliometrics are computed from OpenAlex only (identity source: openalex); citation coverage may lag for very recent 2025-2026 outputs.
- Output and impact dimensions are unscored because the underlying data is missing; h-index 3 and i10 index 1 are the only citation figures provided.
- Only the 2 Evolve-reviewed papers were assessed in depth; the full publication list and publication venues were not independently verified.
- Profile is name-resolved from public bibliographic data (ORCID 0000-0001-7148-0548); no biographical details beyond the listed affiliation are available.