On Provable Copyright Protection for Generative Models

Speaker: Nikhil Vyas , Harvard University

Date: Wednesday, October 18, 2023

Time: 4:00 PM to 5:00 PM Note: all times are in the Eastern Time Zone

Public: Yes

Location: 32-G575

Event Type: Seminar

Room Description: 32-G575

Host: Noah Golowich, MIT

Contact: Noah Golowich, nzg@csail.mit.edu

Relevant URL:

Speaker URL: https://nikhilvyas.github.io/

Speaker Photo:
None

Reminders to: theory-seminars@csail.mit.edu, seminars@csail.mit.edu

Reminder Subject: TALK: Nikhil Vyas: On Provable Copyright Protection for Generative Models

Abstract: There is a growing concern that learned conditional generative models may output samples that are substantially similar to some copyrighted data C that was in their training set. We give a formal definition of near access-freeness (NAF) and prove bounds on the probability that a model satisfying this definition outputs a sample similar to C, even if C is included in its training set. Roughly speaking, a generative model p is k-NAF if for every potentially copyrighted data C, the output of p diverges by at most k-bits from the output of a model q that did not access C at all. We also give generative model learning algorithms, which efficiently modify the original generative model learning algorithm in a black box manner, that output generative models with strong bounds on the probability of sampling protected content. Furthermore, we provide promising experiments showing minimal degradation in output quality while ensuring strong protections against sampling protected content.

Joint work with Sham Kakade and Boaz Barak (https://arxiv.org/abs/2302.10870)

Research Areas:
Algorithms & Theory, AI & Machine Learning

Impact Areas:
Big Data

See other events that are part of the Algorithms and Complexity Seminar 2023.

Created by Noah Golowich at Monday, October 16, 2023 at 12:36 PM.