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

Download:

Current browse context:

cs.CL

Change to browse by:

cs

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 > Computation and Language

Title: BeliefBank: Adding Memory to a Pre-Trained Language Model for a Systematic Notion of Belief

Abstract: Although pretrained language models (PTLMs) contain significant amounts of world knowledge, they can still produce inconsistent answers to questions when probed, even after specialized training. As a result, it can be hard to identify what the model actually "believes" about the world, making it susceptible to inconsistent behavior and simple errors. Our goal is to reduce these problems. Our approach is to embed a PTLM in a broader system that also includes an evolving, symbolic memory of beliefs -- a BeliefBank -- that records but then may modify the raw PTLM answers. We describe two mechanisms to improve belief consistency in the overall system. First, a reasoning component -- a weighted MaxSAT solver -- revises beliefs that significantly clash with others. Second, a feedback component issues future queries to the PTLM using known beliefs as context. We show that, in a controlled experimental setting, these two mechanisms result in more consistent beliefs in the overall system, improving both the accuracy and consistency of its answers over time. This is significant as it is a first step towards PTLM-based architectures with a systematic notion of belief, enabling them to construct a more coherent picture of the world, and improve over time without model retraining.
Comments: EMNLP 2021 Camera Ready. arXiv admin note: substantial text overlap with arXiv:2104.08401
Subjects: Computation and Language (cs.CL)
Cite as: arXiv:2109.14723 [cs.CL]
  (or arXiv:2109.14723v1 [cs.CL] for this version)

Submission history

From: Nora Kassner [view email]
[v1] Wed, 29 Sep 2021 21:04:27 GMT (5747kb,D)

Link back to: arXiv, form interface, contact.