Kunal Talwar
where else to find them
no profile links on file
what they talk about
all 4 in the filters
where they keep going back
no series more than once — all 5 in the filters
on a stage with
- Patrick Foleyonce
- Sean Augensteinonce
- Sewon Minonce
all 4 in the filters
who talks about the same things
nobody close enough yet
| title | conference | date |
|---|---|---|
| Efficient Robust Private Aggregation for DP | Trust in Decentralized Systems 2026 | 2026-03-16 |
Two-server Prio model for secure aggregation, with concrete constructions: PINE: sqrt(D) additional communication, about 3% overhead at D=1,000,000, gives… | ||
| Short Talks | Federated and Collaborative Learning Boot Camp 2026 | 2026-01-26 |
Open-source NVFlare is production-ready since 2021 and runs the 20-hospital COVID oxygen predictor, supports on‑prem IP protection, and surfaces in 50–60+… | ||
| Private Meta-Selection | Privacy and the Science of Data Analysis 2019 | 2019-04-08 |
Defines private meta‑selection: pick the best among K differentially private candidate mechanisms while avoiding an ε cost linear in K. Gives a randomized… | ||
| Large-Scale Private Learning, Part I | Data Privacy: Foundations and Applications Boot Camp | 2019-01-28 |
Covers concrete mechanisms and bounds: the Gaussian mechanism adds N(0,σ^2) with σ = Θ(√(ln(1/δ))/ε) to any ℓ2‑sensitive query, privacy tracked via the… | ||
| Large-Scale Private Learning, Part II | Data Privacy: Foundations and Applications Boot Camp | 2019-01-28 |
Noisy SGD uses per-example gradient clipping and Gaussian noise with privacy amplification by sampling to get utility near nonprivate models; theoretical… | ||
| Differential Privacy | Uncertainty in Computation 2016 | 2016-10-04 |
Survey of real attacks shows reidentification from anonymized records (Massachusetts medical release, Netflix Prize, GWAS/Homer). Formal guarantee treats the… | ||
| nothing matches | ||