https://arxiv.org/abs/2304.03262

close this message

arXiv Forum: How do we make accessible research papers a reality?

If research isn't accessible, can we really call it "Open" Science?
In response to the high interest in this event we have expanded our
online hosting capacity and re-opened registration. Grab your seat.

Learn More
Skip to main content
Cornell University
We gratefully acknowledge support from
the Simons Foundation and member institutions.
 
arxiv logo > cs > arXiv:2304.03262
[                    ]

Help | Advanced Search

[All fields        ]
Search
arXiv logo
Cornell University Logo
[                    ] GO
quick links

  * Login
  * Help Pages
  * About

Computer Science > Artificial Intelligence

arXiv:2304.03262 (cs)
[Submitted on 6 Apr 2023]

Title:When do you need Chain-of-Thought Prompting for ChatGPT?

Authors:Jiuhai Chen, Lichang Chen, Heng Huang, Tianyi Zhou
Download a PDF of the paper titled When do you need Chain-of-Thought
Prompting for ChatGPT?, by Jiuhai Chen and 3 other authors
Download PDF

    Abstract: Chain-of-Thought (CoT) prompting can effectively elicit
    complex multi-step reasoning from Large Language Models~(LLMs).
    For example, by simply adding CoT instruction ``Let's think
    step-by-step'' to each input query of MultiArith dataset, GPT-3's
    accuracy can be improved from 17.7\% to 78.7\%. However, it is
    not clear whether CoT is still effective on more recent
    instruction finetuned (IFT) LLMs such as ChatGPT. Surprisingly,
    on ChatGPT, CoT is no longer effective for certain tasks such as
    arithmetic reasoning while still keeping effective on other
    reasoning tasks. Moreover, on the former tasks, ChatGPT usually
    achieves the best performance and can generate CoT even without
    being instructed to do so. Hence, it is plausible that ChatGPT
    has already been trained on these tasks with CoT and thus
    memorized the instruction so it implicitly follows such an
    instruction when applied to the same queries, even without CoT.
    Our analysis reflects a potential risk of overfitting/bias toward
    instructions introduced in IFT, which becomes more common in
    training LLMs. In addition, it indicates possible leakage of the
    pretraining recipe, e.g., one can verify whether a dataset and
    instruction were used in training ChatGPT. Our experiments report
    new baseline results of ChatGPT on a variety of reasoning tasks
    and shed novel insights into LLM's profiling, instruction
    memorization, and pretraining dataset leakage.

Subjects: Artificial Intelligence (cs.AI)
Cite as:  arXiv:2304.03262 [cs.AI]
          (or arXiv:2304.03262v1 [cs.AI] for this version)
          https://doi.org/10.48550/arXiv.2304.03262
          Focus to learn more
          arXiv-issued DOI via DataCite

Submission history

From: Jiuhai Chen [view email]
[v1] Thu, 6 Apr 2023 17:47:29 UTC (1,275 KB)
Full-text links:

Download:

    Download a PDF of the paper titled When do you need
    Chain-of-Thought Prompting for ChatGPT?, by Jiuhai Chen and 3
    other authors
  * PDF
  * Other formats

[by-nc-sa-4]
Current browse context:
cs.AI
< prev   |   next >
new | recent | 2304
Change to browse by:
cs

References & Citations

  * NASA ADS
  * Google Scholar
  * Semantic Scholar

a export BibTeX citation Loading...

BibTeX formatted citation

x
[loading...          ]
Data provided by:

Bookmark

BibSonomy logo Mendeley logo Reddit logo ScienceWISE logo
(*) Bibliographic Tools

Bibliographic and Citation Tools

[ ] Bibliographic Explorer Toggle
Bibliographic Explorer (What is the Explorer?)
[ ] Litmaps Toggle
Litmaps (What is Litmaps?)
[ ] scite.ai Toggle
scite Smart Citations (What are Smart Citations?)
( ) Code, Data, Media

Code, Data and Media Associated with this Article

[ ] DagsHub Toggle
DagsHub (What is DagsHub?)
[ ] Links to Code Toggle
Papers with Code (What is Papers with Code?)
[ ] ScienceCast Toggle
ScienceCast (What is ScienceCast?)
( ) Demos

Demos

[ ] Replicate Toggle
Replicate (What is Replicate?)
[ ] Spaces Toggle
Hugging Face Spaces (What is Spaces?)
( ) Related Papers

Recommenders and Search Tools

[ ] Link to Influence Flowers
Influence Flowers (What are Influence Flowers?)
[ ] Connected Papers Toggle
Connected Papers (What is Connected Papers?)
[ ] Core recommender toggle
CORE Recommender (What is CORE?)

  * Author
  * Venue
  * Institution
  * Topic

( ) About arXivLabs

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and
share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have
embraced and accepted our values of openness, community, excellence,
and user data privacy. arXiv is committed to these values and only
works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community?
Learn more about arXivLabs.

Which authors of this paper are endorsers? | Disable MathJax (What is
MathJax?)

  * About
  * Help

  * Click here to contact arXiv Contact
  * Click here to subscribe Subscribe

  * Copyright
  * Privacy Policy

  * Web Accessibility Assistance
  * arXiv Operational Status
    Get status notifications via email or slack