Google theorist linking algorithmic game theory and auctions to modern online-learning regret analysis
2
papers here
2
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
24
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—
Not yet scored. The output axis measures production volume and consistency — total works, per-year counts, venue quality. Full publication-level analytics are pending; only raw counts-by-year data and two Evolve-reviewed papers are currently on file.
Impact—
Not yet scored. The impact axis measures citation-based influence: citation counts, velocity, and downstream adoption. H-index 24 and i10-index 39 are recorded, but the fuller citation analytics behind this score are not yet computed, so no estimate is offered.
Focus0/100
Focus scores 0, indicating extreme breadth rather than concentration: the field distribution across the two Evolve-reviewed papers is split exactly evenly (0.5 online learning / learning theory, 0.5 online learning and optimization theory), while the OpenAlex topic profile spans five sizable clusters — game theory and voting systems (76), auction theory and applications (57), optimization and search problems (46), game theory and applications (45), and complexity and algorithms in graphs (42).
Collaboration50/100
Collaboration scores 50, based on 11 distinct coauthors in the OpenAlex window and 32 works over the last three years, implying steady small-team output. The provided frequentCoauthor list signals repeat ties — Yang Cai and Di Wang each appear twice — alongside Vineet Gupta, Yanchen Jiang, Christopher Liaw, Grigoris Velegkas, Anupam Gupta, Guru Guruganesh, and Renato Paes Leme, a pattern typical of theory work.
Activity100/100
Activity scores 100, the axis maximum: recent3yWorks is 32, and countsByYear shows 13 works in 2024 and 14 already recorded in 2026, against only 5 in 2025 and 1-3 per year across 2015-2022. Current intensity matches or exceeds the early-career peaks of 11 works in both 2005 and 2008.
Reproducibility100/100
Reproducibility scores 100, from reproMean 100 over reproPapers 2: both Evolve-reviewed papers carry high repro-readiness. Each is a self-contained theory result — an anytime expert-advice regret bound matching the fixed-time constant, and a deterministic O(d)-regret online inverse optimization proof — and Evolve's human verification found both correct, noting only that the first's overhead vanishes very slowly.
Research summary
Aranyak Mehta is a US-based researcher at Google; this profile is name-resolved from public OpenAlex bibliographic data. The snapshot records an h-index of 24 and an i10-index of 39, with a counts-by-year series starting in 2000 and totaling more than 120 works: peaks of 11 in 2005 and 2008, a long dip to 1-3 per year across 2015-2022, then 13 works in 2024 and 14 already in 2026, for 32 works in the last three years. Topically, the largest OpenAlex clusters are game theory and voting systems (76 works), auction theory and applications (57), optimization and search problems (46), game theory and applications (45), and complexity and algorithms in graphs (42) — a record centred on algorithmic game theory, mechanism design, and algorithms. The two papers Evolve has reviewed, both 2026 arXiv preprints with middle authorship, sit squarely in online learning theory: an anytime prediction-with-expert-advice algorithm whose regret matches the fixed-time constant, and a deterministic O(d)-regret bound for online inverse optimization via trace-power potentials. Both carry high repro-readiness and passed human-verified proof checks, yielding a perfect reproducibility score (reproMean 100 over 2 papers). The field split across reviewed work is even (0.5/0.5), and collaboration is moderate at 11 distinct coauthors, with repeat partners including Vineet Gupta, Christopher Liaw, Grigoris Velegkas, Yang Cai, Di Wang, Anupam Gupta, Guru Guruganesh, and Renato Paes Leme.
Collaboration pattern
Collaboration is moderate (score 50): 11 distinct coauthors appear in the recent OpenAlex window, and 32 works over the last three years suggest steady small-team output. The provided frequentCoauthor list shows repeat ties — Yang Cai and Di Wang each appear twice — alongside Vineet Gupta, Yanchen Jiang, Christopher Liaw, Grigoris Velegkas, Anupam Gupta, Guru Guruganesh, and Renato Paes Leme, a mix of industry colleagues and academic collaborators typical of theory research.
Research keywords
Highlights
Anytime expert-advice regret matches the fixed-time constant
Deterministic O(d) regret for online inverse optimization
Late-career publication surge in the OpenAlex record
After 1-3 works per year from 2015 to 2022, the counts-by-year series jumps to 13 works in 2024 and 14 already recorded in 2026, giving 32 works over the last three years — the most intensive stretch in the snapshot.
Timeline
- 2000First year in the provided counts-by-year series: 1 work recorded, followed by 1 in 2002 and 5 in 2003.
- 2005Early peak: 11 works recorded — the highest annual count until 2024.
- 2008Second peak of the late-2000s run: 11 works again, after 8 in 2007 and 6 in 2006.
- 2015Output thins: 1 work in 2015, repeating in 2017 and 2018, with 1-3 per year through 2022.
- 2024Renewed surge: 13 works recorded, the highest single-year count in the snapshot.
- 202614 works recorded to date; both Evolve-reviewed arXiv papers fall here — anytime regret for prediction with expert advice, and O(d)-regret online inverse optimization.
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 (works, topics, coauthors, counts by year) come from an OpenAlex-based snapshot; venue and citation coverage may be incomplete.
- H-index 24 and i10-index 39 are quoted verbatim from the provided raw metrics; the fuller citation analytics needed for output/impact scoring are not yet computed.
- Evolve verdicts cover only the two 2026 arXiv papers reviewed; no other works in the record have been human-verified.
- The 2026 counts-by-year figure (14) reflects a partial year, and the provided frequentCoauthor list contains duplicate entries (Yang Cai, Di Wang).