2026-09-19T00:20:04Zhttps://dspace.mit.edu/server/oai/requestoai:dspace.mit.edu:1721.1/1242522026-06-06T00:49:00Zcom_1721.1_7582com_1721.1_7581col_1721.1_131023
David Sontag.
Khandelwal, Arjun(Arjun Sunil)
Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science.
Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science
2020-03-24T15:36:29Z
2020-03-24T15:36:29Z
2019
2019
https://hdl.handle.net/1721.1/124252
1145123030
This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections.
Thesis: M. Eng., Massachusetts Institute of Technology, Department of Electrical Engineering and Computer Science, 2019
Cataloged from student-submitted PDF version of thesis.
Includes bibliographical references (pages 73-76).
Deep generative models have emerged as a powerful modeling paradigm for making sense of large amounts of unlabeled real-world data. In particular, the representations produced by these models have proven to be useful both in improving human understanding of the factors of variation in the original dataset and in downstream tasks such as classification. Most current algorithms, however, require training a bespoke model from scratch, which can be both expensive and time-consuming. Instead, we propose various methods of fine-tuning pre-trained generative models to achieve these goals, and evaluate these methods quantitatively on few-shot classification and interpretability tasks.
by Arjun Khandelwal.
M.Eng.
M.Eng. Massachusetts Institute of Technology, Department of Electrical Engineering and Computer Science
76 pages
eng
Massachusetts Institute of Technology
MIT theses are protected by copyright. They may be viewed, downloaded, or printed from this source but further reproduction or distribution in any format is prohibited without written permission.
http://dspace.mit.edu/handle/1721.1/7582
Electrical Engineering and Computer Science.
Fine-tuning generative models
Thesis
application/pdf
2020-03-24T15:36:29Z
Publication
Master
EECS
unknown
unknown
<Publication xmlns="https://www.openaire.eu/cerif-profile/1.1/" id="25bf58e2-a92c-433e-8d62-1ae4b5ce779c">
<Type xmlns="https://www.openaire.eu/cerif-profile/vocab/COAR_Publication_Types">http://purl.org/coar/resource_type/c_1843</Type>
<Language>eng</Language>
<Title>Fine-tuning generative models</Title>
<PublishedIn>
<Publication>
</Publication>
</PublishedIn>
<PublicationDate>2019</PublicationDate>
<Authors>
<Author>
<DisplayName>Khandelwal, Arjun(Arjun Sunil)</DisplayName>
<Affiliation>
<OrgUnit>
</OrgUnit>
</Affiliation>
</Author>
</Authors>
<Editors>
</Editors>
<Publishers>
<Publisher>
<DisplayName>Massachusetts Institute of Technology</DisplayName>
<OrgUnit />
</Publisher>
</Publishers>
<License>http://dspace.mit.edu/handle/1721.1/7582</License>
<Keyword>Electrical Engineering and Computer Science.</Keyword>
<Abstract>Deep generative models have emerged as a powerful modeling paradigm for making sense of large amounts of unlabeled real-world data. In particular, the representations produced by these models have proven to be useful both in improving human understanding of the factors of variation in the original dataset and in downstream tasks such as classification. Most current algorithms, however, require training a bespoke model from scratch, which can be both expensive and time-consuming. Instead, we propose various methods of fine-tuning pre-trained generative models to achieve these goals, and evaluate these methods quantitatively on few-shot classification and interpretability tasks.</Abstract>
<Access xmlns="http://purl.org/coar/access_right"
>
</Access>
</Publication>