Skip to main content
Cornell University
We gratefully acknowledge support from the Simons Foundation, member institutions, and all contributors. Donate
arxiv logo > cs > arXiv:2301.11270

Help | Advanced Search

arXiv logo
Cornell University Logo

quick links

  • Login
  • Help Pages
  • About

Computer Science > Machine Learning

arXiv:2301.11270 (cs)
[Submitted on 26 Jan 2023 (v1), last revised 8 Feb 2024 (this version, v5)]

Title:Principled Reinforcement Learning with Human Feedback from Pairwise or $K$-wise Comparisons

Authors:Banghua Zhu, Jiantao Jiao, Michael I. Jordan
View a PDF of the paper titled Principled Reinforcement Learning with Human Feedback from Pairwise or $K$-wise Comparisons, by Banghua Zhu and 2 other authors
View PDF HTML (experimental)
Abstract:We provide a theoretical framework for Reinforcement Learning with Human Feedback (RLHF). Our analysis shows that when the true reward function is linear, the widely used maximum likelihood estimator (MLE) converges under both the Bradley-Terry-Luce (BTL) model and the Plackett-Luce (PL) model. However, we show that when training a policy based on the learned reward model, MLE fails while a pessimistic MLE provides policies with improved performance under certain coverage assumptions. Additionally, we demonstrate that under the PL model, the true MLE and an alternative MLE that splits the $K$-wise comparison into pairwise comparisons both converge. Moreover, the true MLE is asymptotically more efficient. Our results validate the empirical success of existing RLHF algorithms in InstructGPT and provide new insights for algorithm design. Furthermore, our results unify the problem of RLHF and max-entropy Inverse Reinforcement Learning (IRL), and provide the first sample complexity bound for max-entropy IRL.
Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC); Statistics Theory (math.ST); Machine Learning (stat.ML)
Cite as: arXiv:2301.11270 [cs.LG]
  (or arXiv:2301.11270v5 [cs.LG] for this version)
  https://doi.org/10.48550/arXiv.2301.11270
arXiv-issued DOI via DataCite

Submission history

From: Banghua Zhu [view email]
[v1] Thu, 26 Jan 2023 18:07:21 UTC (114 KB)
[v2] Mon, 30 Jan 2023 00:10:09 UTC (121 KB)
[v3] Sun, 19 Mar 2023 06:32:21 UTC (121 KB)
[v4] Thu, 1 Jun 2023 01:05:21 UTC (122 KB)
[v5] Thu, 8 Feb 2024 04:16:52 UTC (122 KB)
Full-text links:

Access Paper:

    View a PDF of the paper titled Principled Reinforcement Learning with Human Feedback from Pairwise or $K$-wise Comparisons, by Banghua Zhu and 2 other authors
  • View PDF
  • HTML (experimental)
  • TeX Source
  • Other Formats
view license
Current browse context:
cs.LG
< prev   |   next >
new | recent | 2023-01
Change to browse by:
cs
cs.AI
cs.HC
math
math.ST
stat
stat.ML
stat.TH

References & Citations

  • NASA ADS
  • Google Scholar
  • Semantic Scholar
a export BibTeX citation Loading...

BibTeX formatted citation

×
Data provided by:

Bookmark

BibSonomy logo Reddit logo

Bibliographic and Citation Tools

Bibliographic Explorer (What is the Explorer?)
Connected Papers (What is Connected Papers?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)

Code, Data and Media Associated with this Article

alphaXiv (What is alphaXiv?)
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Hugging Face (What is Huggingface?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)

Demos

Replicate (What is Replicate?)
Hugging Face Spaces (What is Spaces?)
TXYZ.AI (What is TXYZ.AI?)

Recommenders and Search Tools

Influence Flower (What are Influence Flowers?)
CORE Recommender (What is CORE?)
IArxiv Recommender (What is IArxiv?)
  • Author
  • Venue
  • Institution
  • Topic

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
  • contact arXivClick here to contact arXiv Contact
  • subscribe to arXiv mailingsClick here to subscribe Subscribe
  • Copyright
  • Privacy Policy
  • Web Accessibility Assistance
  • arXiv Operational Status
    Get status notifications via email or slack