We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Machine Learning

Title: Controlling Conditional Language Models with Distributional Policy Gradients

Abstract: Machine learning is shifting towards general-purpose pretrained generative models, trained in a self-supervised manner on large amounts of data, which can then be applied to solve a large number of tasks. However, due to their generic training methodology, these models often fail to meet some of the downstream requirements (e.g. hallucination in abstractive summarization or wrong format in automatic code generation). This raises an important question on how to adapt pre-trained generative models to a new task without destroying its capabilities. Recent work has suggested to solve this problem by representing task-specific requirements through energy-based models (EBMs) and approximating these EBMs using distributional policy gradients (DPG). Unfortunately, this approach is limited to unconditional distributions, represented by unconditional EBMs. In this paper, we extend this approach to conditional tasks by proposing Conditional DPG (CDPG). We evaluate CDPG on three different control objectives across two tasks: summarization with T5 and code generation with GPT-Neo. Our results show that fine-tuning using CDPG robustly moves these pretrained models closer towards meeting control objectives and -- in contrast with baseline approaches -- does not result in catastrophic forgetting.
Comments: CtrlGen: Controllable Generative Modeling in Language and Vision Workshop at NeurIPS 2021
Subjects: Machine Learning (cs.LG); Computation and Language (cs.CL)
Cite as: arXiv:2112.00791 [cs.LG]
  (or arXiv:2112.00791v1 [cs.LG] for this version)

Submission history

From: Tomek Korbak [view email]
[v1] Wed, 1 Dec 2021 19:24:05 GMT (304kb,D)

Link back to: arXiv, form interface, contact.