Trust in Decentralized Systems

Program

Federated and Collaborative Learning

Location

Calvin Lab Auditorium

Date

Monday, Mar. 16 – Friday, Mar. 20, 2026

Back to calendar

All talks listed in Pacific Time. Schedule subject to change.

9:15 – 9:45 a.m.

Coffee and Check-In

9:45 – 10:30 a.m.

Privacy Amplification by Sampling in Practice

Arun Ganesh (Google)

10:30 – 11 a.m.

Break

11 – 11:30 a.m.

Privacy amplification by random allocation

Vitaly Feldman (Apple)

11:30 a.m. – 12 p.m.

Heterogeneity and Privacy in Modern Learning

Maryam Aliakbarpour (Rice University)

12 – 1:30 p.m.

Lunch (on your own)

2 – 2:30 p.m.

Chasing the Constants and Its Implication in Private Learning

Jalaj Upadhyay (Rutgers University)

2:30 – 3 p.m.

Privately Estimating Black-Box Statistics

Thomas Steinke (Google DeepMind)

3 – 3:30 p.m.

"Having Confidence in My Confidence Intervals": How Data Users Engage with Privacy-Protected Wikipedia Data

Jayshree Sarathy (Northeastern)

3:30 – 4 p.m.

Pairwise Network Differential Privacy

Edwige Cyffers (CNRS)

4 – 5 p.m.

Reception

9:15 – 9:45 a.m.

Coffee and Check-In

9:45 – 10:30 a.m.

When Unstoppable Frontier Agents Meet Immovable Attack Vectors

Eugene Badgasaryan (UMass)

10:30 – 11 a.m.

Break

11 – 11:30 a.m.

2026 Is the New 2016, but Make It Privacy: On Federated Memory, Contextual Privacy, and Personalized Agents

Niloofar Mireshghallah (Carnegie Mellon University)

11:30 a.m. – 12 p.m.

VaultGemma: A Differentially Private Gemma Model

Amer Sinha (Google)

12 – 2 p.m.

Lunch (on your own)

2 – 2:30 p.m.

Contextual Privacy in the Agentic Era (Virtual Talk)

Janardhan Kulkarni (Microsoft Research)

2:30 – 3 p.m.

Private Insights into AI Use

Peter Kairouz (Google AI)

3 – 3:30 p.m.

Break

3:30 – 4:15 p.m.

Synthetic Data as an Enabler for Learning from Decentralized Private Data

Giulia Fanti (Carnegie Mellon University)

4:15 – 5 p.m.

Privacy in Practice: Architecting Differential Privacy into Web Advertising Standards

Roxana Geambasu (Columbia University)

9:15 – 9:45 a.m.

Coffee and Check-In

9:45 – 10:30 a.m.

Seamless auditing privacy

Saeed Mahloujifar (Meta)

10:30 – 11 a.m.

Break

11 – 11:30 a.m.

The Sample Complexity of Membership Inference and Privacy Auditing (Virtual Talk)

Jonathan Ullman (Northeastern)

11:30 a.m. – 12 p.m.

GeoClip: Geometry-Aware Clipping for Differentially Private SGD

Lalitha Sankar (Arizona State University)

12 – 2 p.m.

Lunch (on your own)

2 – 2:30 p.m.

Stargazing into AI Usage with Differential Privacy

Pritish Kamath (Google)

2:30 – 3 p.m.

Talk by

Kunal Talwar (Apple Inc)

3 – 3:30 p.m.

Break

3:30 – 4 p.m.

Distributed Models for Private Analysis of Graph Data

Adam Smith (Boston University)

4 – 4:30 p.m.

From Theory to Practice: Advances in Real-World Federated Learning with NVIDIA FLARE

Holger Roth (Nvidia)

4:30 – 5 p.m.

Dimension-free Private Mean Estimation for Anisotropic Distributions

Lydia Zakynthinou (Johns Hopkins University)

5 – 6 p.m.

Hike

9 – 9:30 a.m.

Coffee and Check-In

9:30 – 10:30 a.m.

WIP title: Building Practical Privacy-Preserving Inference Systems (Virtual Talk)

Wenting Zheng (Carnegie Mellon University)

10:30 – 11 a.m.

Break

11 – 11:30 a.m.

Towards making private telemetry as ubiquitous as TLS

Ryan Lehmkuhl (UC Berkeley)

11 – 11:30 a.m.

Towards making private telemetry as ubiquitous as TLS

Ryan Lehmkuhl (UC Berkeley)

11:30 a.m. – 12 p.m.

Private aggregation at scale with decentralized trust

Yiping Ma (UC Berkeley)

12 – 2 p.m.

Lunch (on your own)

2 – 2:30 p.m.

Secure Aggregation with Lightweight Committees

Phillipp Schoppmann (Google)

2:30 – 3 p.m.

Distributed Aggregation Protocol: A Standard for Privacy-Preserving Aggregation in MPC

Tim Geoghegan (Internet Security Research Group)

3 – 3:30 p.m.

Break

3:30 – 4 p.m.

Splitting Secrets for Encrypted Backups

Emma Dauterman (Stanford)

4 – 4:30 p.m.

Talk by

Rehan Rishi (Apple)

4:30 – 5 p.m.

AgentCrypt: Advancing Privacy and (Secure) Computation in AI Agent Collaboration

Antigoni Polychroniadou (J.P. Morgan AI Research)

9 – 9:30 a.m.

Coffee and Check-In

9:30 – 10:30 a.m.

Verifiable Data Science

Guy Rothblum (Apple)

10:30 – 11 a.m.

Break

11 – 11:30 a.m.

Hardening Confidential Federated Computations against Side-Channel Attacks (Virtual Talk)

Albert Cheu (Google)

11:30 a.m. – 12 p.m.

Personalized Federated Diffusion models and online learning with connections to privacy

Suhas Diggavi (UCLA)

12 – 2 p.m.

Lunch (on your own)

2 – 2:30 p.m.

Local Pan-Privacy for Federated Analytics

Audra McMillan (Apple Inc)

2:30 – 3 p.m.

Training generative models from locally privatized data via entropic optimal transport

Ayfer Ozgur (Stanford University)

3 – 3:30 p.m.

Break

3:30 – 4 p.m.

Private Geometric Median

Mahdi Haghifam (Toyota Technological Institute at Chicago)

4 – 4:30 p.m.

Personalized Federated Training of Diffusion Models with Privacy Guarantees

Kumar Kshitij Patel (Yale University)