Journal article · 2026 · Open access

Partial identifiability and misspecification in inverse reinforcement learning

Joar Skalse, Alessandro Abate

Artificial Intelligence, 356, 104525 · Published 25 March 2026 · DOI 10.1016/j.artint.2026.104525

Open full textCite this paper

In brief

“Partial identifiability and misspecification in inverse reinforcement learning” is a 2026 journal article by Joar Skalse and Alessandro Abate, published in Artificial Intelligence. OpenAlex counts 9 citations as of 17 November 2026, which puts it in the top 5% of works published in 2026 by citations. It is open access under a CC BY licence: the full text is free to read at Artificial Intelligence.

Citations
9
Most cited in
2026 (6)
Field-weighted impact
8.37
References
55
Authors
2

Abstract

The aim of Inverse Reinforcement Learning (IRL) is to infer a reward function R from a policy π . This problem is difficult, for several reasons. First of all, there are typically multiple reward functions which are compatible with a given policy; this means that the reward function is only partially identifiable , and that IRL contains a certain fundamental degree of ambiguity. Secondly, in order to infer R from π , an IRL algorithm must have a behavioural model of how π relates to R . However, the true relationship between human preferences and human behaviour is very complex, and practically impossible to fully capture with a simple model. This means that the behavioural model in practice will be misspecified , which raises the worry that it might lead to unsound inferences if applied to real-world data. In this paper, we provide a comprehensive mathematical analysis of partial identifiability and misspecification in IRL. Specifically, we fully characterise and quantify the ambiguity of the reward function for all of the behavioural models that are most common in the current IRL literature. We also provide necessary and sufficient conditions that describe precisely how the observed demonstrator policy may differ from each of the standard behavioural models before that model leads to faulty inferences about the reward function R . In addition to this, we introduce a cohesive framework for reasoning about partial identifiability and misspecification in IRL, together with several formal tools that can be used to easily derive the partial identifiability and misspecification robustness of new IRL models, or analyse other kinds of reward learning algorithms.

Cite this paper

Formatted from the catalogue record. Check names, capitals and page numbers against the paper before you submit.

APA 7

Skalse, J., & Abate, A. (2026). Partial identifiability and misspecification in inverse reinforcement learning. Artificial Intelligence, 356, 104525. https://doi.org/10.1016/j.artint.2026.104525

MLA 9

Skalse, Joar, and Alessandro Abate. “Partial identifiability and misspecification in inverse reinforcement learning.” Artificial Intelligence, vol. 356, 2026, p. 104525. https://doi.org/10.1016/j.artint.2026.104525.

Chicago

Skalse, Joar and Alessandro Abate. 2026. “Partial identifiability and misspecification in inverse reinforcement learning.” Artificial Intelligence 356: 104525. https://doi.org/10.1016/j.artint.2026.104525.

Harvard

Skalse, J. and Abate, A. (2026) ‘Partial identifiability and misspecification in inverse reinforcement learning’, Artificial Intelligence, 356, p. 104525. doi:10.1016/j.artint.2026.104525.

IEEE

J. Skalse and A. Abate, “Partial identifiability and misspecification in inverse reinforcement learning,” Artificial Intelligence, vol. 356, p. 104525, 2026, doi: 10.1016/j.artint.2026.104525.

BibTeX

@article{skalse2026partial,
  title     = {{Partial identifiability and misspecification in inverse reinforcement learning}},
  author    = {Skalse, Joar and Abate, Alessandro},
  journal   = {Artificial Intelligence},
  year      = {2026},
  volume    = {356},
  pages     = {104525},
  doi       = {10.1016/j.artint.2026.104525},
  url       = {https://doi.org/10.1016/j.artint.2026.104525}
}

Download for a reference manager: BibTeX (.bib) · RIS for Zotero, Mendeley and EndNote (.ris)

Topic and keywords

Physical Sciences › Computer Science › Artificial Intelligence › Reinforcement Learning in Robotics

  • inverse reinforcement learning
  • model misspecification
  • reward function inference
  • behavioural models
  • robustness analysis
  • human preferences
  • reward learning

Work with this paper

Most-cited papers in Computer Science

All highly cited Computer Science papers

Questions about this paper

How do I cite this paper in APA style?

Skalse, J., & Abate, A. (2026). Partial identifiability and misspecification in inverse reinforcement learning. Artificial Intelligence, 356, 104525. https://doi.org/10.1016/j.artint.2026.104525

How many times has this paper been cited?

9 times, according to OpenAlex on 17 November 2026. Google Scholar and Scopus count different sets of sources, so their numbers differ.

Is this paper free to read?

Yes. It is open access at Artificial Intelligence: sign in and open the full text from this page.

Who wrote this paper?

Joar Skalse, Alessandro Abate.

What is the DOI of this paper?

10.1016/j.artint.2026.104525, permanently at https://doi.org/10.1016/j.artint.2026.104525. Sign in and open it from this page.

Bibliographic data from OpenAlex (CC0), last checked 17 November 2026. ResearchGuru does not host this paper; once you are signed in, the links open it at the publisher or the repository that does.