upcarta
  • Sign In
  • Sign Up
  • Explore
  • Search

Do language models possess knowledge (soundness)?

  • Article
  • Apr, 2022
  • #Philosophyofcomputerscience #ProgrammingLanguages #Naturallanguageprocessing #InformationTheory
Tarun Chitra
@tarunchitra
(Author)
Henry Prior
@priorupdates
(Author)
hackmd.io
Read on hackmd.io
1 Recommender
2 Mentions
1 Collection
In this post we explore how interactive proofs, a concept of interest in theoretical computer science and cryptography, can be useful for testing capabilities of language models. We... Show More

In this post we explore how interactive proofs, a concept of interest in theoretical computer science and cryptography, can be useful for testing capabilities of language models. We consider how a setup comprised of multiple language models could be used in place of prompt tuning for this purpose. Our goal is to try to understand the role of knowledge in language models — can they be viewed (at least to one another) as possessing knowledge? We utilize tools from interactive proofs in theoretical computer science to try to formalize this and provide a mathematical lens on the question, “do large language models actually possess knowledge?”

Show Less
Recommend
Post
Save
Complete
Collect
Mentions
See All
Cemre Ucar @CemreUcar · May 17, 2023
  • Post
beyond good. "The beauty of interactive proof systems is that by separating the computational complexity of the program from that of the adversaries, one can formally guarantee knowledge soundness."
Cemre Ucar @CemreUcar · May 24, 2023
  • Curated in Artificial Intelligence Tools and Sustainability Implementation Workshop Resources - Şirince 2023
Collections
See All
  • Cemre Ucar
    • Collection
    Artificial Intelligence Tools and Sustainability Implementation Workshop Resources - Şirince 2023
    5 curations
  • upcarta ©2025
  • Home
  • About
  • Terms
  • Privacy
  • Cookies
  • @upcarta