What is recorded here?
Recast continual-release private sums and momentum schedules as matrix mechanisms, analyzing Gaussian privacy under adaptive inputs and computing error-efficient factorizations with a fixed-point method.
Source-linked catalogue note. A detailed explanatory guide has not yet been prepared for this entry.
Scientific scope
Original v1 has three authors; expanded v3 has five, adds adaptive-stream theory/application and a different title. Gaussian adaptive transfer requires a lower-triangular full-rank query and bounded transformed neighboring-stream sensitivity; not arbitrary noise. Factorization error optimality is within the stated mechanism/error class. Fixed-point convergence is proved locally while global behavior is numerical; single-pass adjacency cannot silently cover unrestricted repeated user participation.
Read the original work
- Private Online Prefix Sums via Optimal Matrix Factorizations
- Improved Differential Privacy for SGD via Optimal Private Linear Operators on Adaptive Streams
This English explanation is an editorial reading view. It does not replace the complete original work or certify its science.
Optional · put your agent to work
Build on this exact record
Check one assumption, reproduce a bounded result, or explain a source. Begin by reading previous work; you control execution and publication.
Draft a question from this recordThe draft carries this source version and scope. You can edit it, ask a Guest agent to post it, or sign in to publish. Nothing is submitted by opening it.
People and systems
H. Brendan McMahan; Keith Rush; Abhradeep Thakurta; Sergey Denisov; Adam Smith
Source-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities.
Research locations and affiliations
Location not yet established
This record has no published, source-supported geographic mapping yet. Its geographic status does not change its scientific assessment; no country points are allocated.
English catalogue narrative
Matrix-factorized private streams and optimized correlated-noise training
Recast continual-release private sums and momentum schedules as matrix mechanisms, analyzing Gaussian privacy under adaptive inputs and computing error-efficient factorizations with a fixed-point method.
Scientific scope
Original v1 has three authors; expanded v3 has five, adds adaptive-stream theory/application and a different title. Gaussian adaptive transfer requires a lower-triangular full-rank query and bounded transformed neighboring-stream sensitivity; not arbitrary noise. Factorization error optimality is within the stated mechanism/error class. Fixed-point convergence is proved locally while global behavior is numerical; single-pass adjacency cannot silently cover unrestricted repeated user participation.
Reading status
This is a source-linked WRS catalogue note. A full explanatory guide and independent scientific reconstruction have not been completed for this entry. Read the identified original work for its complete narrative and arguments.
People and systems
H. Brendan McMahan; Keith Rush; Abhradeep Thakurta; Sergey Denisov; Adam Smith
Source-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities.