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

Download:

Current browse context:

cs.CL
< prev | next >
new | recent | 2301

Change to browse by:

cs
cs.LG

References & Citations

DBLP - CS Bibliography

listing | bibtex

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: Causal-Discovery Performance of ChatGPT in the context of Neuropathic Pain Diagnosis

Authors: Ruibo Tu, Chao Ma, Cheng Zhang
(Submitted on 24 Jan 2023 (v1), last revised 6 Feb 2023 (this version, v2))
Abstract: ChatGPT has demonstrated exceptional proficiency in natural language conversation, e.g., it can answer a wide range of questions while no previous large language models can. Thus, we would like to push its limit and explore its ability to answer causal discovery questions by using a medical benchmark (Tu et al. 2019) in causal discovery.
Subjects: Computation and Language (cs.CL); Machine Learning (cs.LG)
Cite as: arXiv:2301.13819 [cs.CL]
  (or arXiv:2301.13819v2 [cs.CL] for this version)

Submission history

From: Cheng Zhang [view email]
[v1] Tue, 24 Jan 2023 19:23:38 GMT (882kb,D)
[v2] Mon, 6 Feb 2023 12:03:38 GMT (882kb,D)
Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)

Link back to: arXiv, form interface, contact.