[2309.12288] The Reversal Curse: LLMs trained on "A is B" fail to learn "B is A"

-->

Computer Science > Computation and Language

arXiv:2309.12288 (cs)

[Submitted on 21 Sep 2023 (v1), last revised 26 May 2024 (this version, v4)]

Title:The Reversal Curse: LLMs trained on "A is B" fail to learn "B is A"

Authors:Lukas Berglund, Meg Tong, Max Kaufmann, Mikita Balesni, Asa Cooper Stickland, Tomasz Korbak, Owain Evans<br>View a PDF of the paper titled The Reversal Curse: LLMs trained on "A is B" fail to learn "B is A", by Lukas Berglund and 6 other authors

View PDF<br>HTML (experimental)

Abstract:We expose a surprising failure of generalization in auto-regressive large language models (LLMs). If a model is trained on a sentence of the form "A is B", it will not automatically generalize to the reverse direction "B is A". This is the Reversal Curse. For instance, if a model is trained on "Valentina Tereshkova was the first woman to travel to space", it will not automatically be able to answer the question, "Who was the first woman to travel to space?". Moreover, the likelihood of the correct answer ("Valentina Tershkova") will not be higher than for a random name. Thus, models do not generalize a prevalent pattern in their training set: if "A is B" occurs, "B is A" is more likely to occur. It is worth noting, however, that if "A is B" appears in-context, models can deduce the reverse relationship. We provide evidence for the Reversal Curse by finetuning GPT-3 and Llama-1 on fictitious statements such as "Uriah Hawthorne is the composer of Abyssal Melodies" and showing that they fail to correctly answer "Who composed Abyssal Melodies?". The Reversal Curse is robust across model sizes and model families and is not alleviated by data augmentation. We also evaluate ChatGPT (GPT-3.5 and GPT-4) on questions about real-world celebrities, such as "Who is Tom Cruise's mother? [A: Mary Lee Pfeiffer]" and the reverse "Who is Mary Lee Pfeiffer's son?". GPT-4 correctly answers questions like the former 79% of the time, compared to 33% for the latter.

Code available at: this https URL.

Comments:<br>21 pages, 11 figures

Subjects:

Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)

Cite as:<br>arXiv:2309.12288 [cs.CL]

(or<br>arXiv:2309.12288v4 [cs.CL] for this version)

https://doi.org/10.48550/arXiv.2309.12288

Focus to learn more

arXiv-issued DOI via DataCite

Submission history<br>From: Owain Evans [view email]<br>[v1]<br>Thu, 21 Sep 2023 17:52:19 UTC (1,320 KB)

[v2]<br>Fri, 22 Sep 2023 18:08:20 UTC (1,319 KB)

[v3]<br>Thu, 4 Apr 2024 21:25:17 UTC (1,336 KB)

[v4]<br>Sun, 26 May 2024 17:45:21 UTC (1,336 KB)

Full-text links:<br>Access Paper:

View a PDF of the paper titled The Reversal Curse: LLMs trained on "A is B" fail to learn "B is A", by Lukas Berglund and 6 other authors<br>View PDF<br>HTML (experimental)<br>TeX Source

view license

Current browse context:

cs.CL

next >

new<br>recent<br>| 2023-09

Change to browse by:

cs<br>cs.AI<br>cs.LG

References & Citations

NASA ADS<br>Google Scholar

Semantic Scholar

1 blog link<br>(what is this?)

export BibTeX citation<br>Loading...

BibTeX formatted citation

&times;

loading...

Data provided by:

Bookmark

Bibliographic Tools

Bibliographic and Citation Tools

Bibliographic Explorer Toggle

Bibliographic Explorer (What is the Explorer?)

Connected Papers Toggle

Connected Papers (What is Connected Papers?)

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

alphaXiv Toggle

alphaXiv (What is alphaXiv?)

Links to Code Toggle

CatalyzeX Code Finder for Papers (What is CatalyzeX?)

DagsHub Toggle

DagsHub (What is DagsHub?)

GotitPub Toggle

Gotit.pub (What is GotitPub?)

Huggingface Toggle

Hugging Face (What is Huggingface?)

ScienceCast Toggle

ScienceCast (What is ScienceCast?)

Demos

Demos

Replicate Toggle

Replicate (What is Replicate?)

Spaces Toggle

Hugging Face Spaces (What is Spaces?)

Spaces Toggle

TXYZ.AI (What is TXYZ.AI?)

Related Papers

Recommenders and Search Tools

Link to Influence Flower

Influence Flower (What are Influence Flowers?)

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? |<br>Disable MathJax (What is MathJax?)