@inproceedings{nwadike-etal-2025-library,
title = "Library-Like Behavior In Language Models is Enhanced by Self-Referencing Causal Cycles",
author = "Nwadike, Munachiso S and
Iklassov, Zangir and
Aremu, Toluwani and
Hiraoka, Tatsuya and
Heinzerling, Benjamin and
Bojkovic, Velibor and
AlQuabeh, Hilal and
Tak{\'a}{\v{c}}, Martin and
Inui, Kentaro",
editor = "Che, Wanxiang and
Nabende, Joyce and
Shutova, Ekaterina and
Pilehvar, Mohammad Taher",
booktitle = "Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)",
month = jul,
year = "2025",
address = "Vienna, Austria",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.acl-long.1232/",
doi = "10.18653/v1/2025.acl-long.1232",
pages = "25365--25377",
ISBN = "979-8-89176-251-0",
abstract = "We introduce the concept of the $\textit{self-referencing causal cycle}$ (abbreviated $\textit{ReCall}$ ){---}a mechanism that enables large language models (LLMs) to bypass the limitations of unidirectional causality, which underlies a phenomenon known as the $\textit{reversal curse}$. When an LLM is prompted with sequential data, it often fails to recall preceding context. For example, when we ask an LLM to recall the line preceding ``O say does that star-spangled banner yet wave'' in the U.S. National Anthem, it often fails to correctly return ``Gave proof through the night that our flag was still there''{---}this is due to the reversal curse. It occurs because language models such as ChatGPT and Llama generate text based on preceding tokens, requiring facts to be learned and reproduced in a consistent token order. While the reversal curse is often viewed as a limitation, we offer evidence of an alternative view: it is not always an obstacle in practice. We find that $\textit{ReCall}$ is driven by what we designate as $\textit{cycle tokens}${---}sequences that connect different parts of the training data, enabling recall of preceding tokens from succeeding ones. Through rigorous probabilistic formalization and controlled experiments, we demonstrate how the cycles they induce influence a model{'}s ability to reproduce information. To facilitate reproducibility, we provide our code and experimental details at https://anonymous.4open.science/r/remember-B0B8/."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="nwadike-etal-2025-library">
<titleInfo>
<title>Library-Like Behavior In Language Models is Enhanced by Self-Referencing Causal Cycles</title>
</titleInfo>
<name type="personal">
<namePart type="given">Munachiso</namePart>
<namePart type="given">S</namePart>
<namePart type="family">Nwadike</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zangir</namePart>
<namePart type="family">Iklassov</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Toluwani</namePart>
<namePart type="family">Aremu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tatsuya</namePart>
<namePart type="family">Hiraoka</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Benjamin</namePart>
<namePart type="family">Heinzerling</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Velibor</namePart>
<namePart type="family">Bojkovic</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hilal</namePart>
<namePart type="family">AlQuabeh</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Martin</namePart>
<namePart type="family">Takáč</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kentaro</namePart>
<namePart type="family">Inui</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Wanxiang</namePart>
<namePart type="family">Che</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Joyce</namePart>
<namePart type="family">Nabende</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ekaterina</namePart>
<namePart type="family">Shutova</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mohammad</namePart>
<namePart type="given">Taher</namePart>
<namePart type="family">Pilehvar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Vienna, Austria</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-251-0</identifier>
</relatedItem>
<abstract>We introduce the concept of the self-referencing causal cycle (abbreviated ReCall )—a mechanism that enables large language models (LLMs) to bypass the limitations of unidirectional causality, which underlies a phenomenon known as the reversal curse. When an LLM is prompted with sequential data, it often fails to recall preceding context. For example, when we ask an LLM to recall the line preceding “O say does that star-spangled banner yet wave” in the U.S. National Anthem, it often fails to correctly return “Gave proof through the night that our flag was still there”—this is due to the reversal curse. It occurs because language models such as ChatGPT and Llama generate text based on preceding tokens, requiring facts to be learned and reproduced in a consistent token order. While the reversal curse is often viewed as a limitation, we offer evidence of an alternative view: it is not always an obstacle in practice. We find that ReCall is driven by what we designate as cycle tokens—sequences that connect different parts of the training data, enabling recall of preceding tokens from succeeding ones. Through rigorous probabilistic formalization and controlled experiments, we demonstrate how the cycles they induce influence a model’s ability to reproduce information. To facilitate reproducibility, we provide our code and experimental details at https://anonymous.4open.science/r/remember-B0B8/.</abstract>
<identifier type="citekey">nwadike-etal-2025-library</identifier>
<identifier type="doi">10.18653/v1/2025.acl-long.1232</identifier>
<location>
<url>https://aclanthology.org/2025.acl-long.1232/</url>
</location>
<part>
<date>2025-07</date>
<extent unit="page">
<start>25365</start>
<end>25377</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Library-Like Behavior In Language Models is Enhanced by Self-Referencing Causal Cycles
%A Nwadike, Munachiso S.
%A Iklassov, Zangir
%A Aremu, Toluwani
%A Hiraoka, Tatsuya
%A Heinzerling, Benjamin
%A Bojkovic, Velibor
%A AlQuabeh, Hilal
%A Takáč, Martin
%A Inui, Kentaro
%Y Che, Wanxiang
%Y Nabende, Joyce
%Y Shutova, Ekaterina
%Y Pilehvar, Mohammad Taher
%S Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
%D 2025
%8 July
%I Association for Computational Linguistics
%C Vienna, Austria
%@ 979-8-89176-251-0
%F nwadike-etal-2025-library
%X We introduce the concept of the self-referencing causal cycle (abbreviated ReCall )—a mechanism that enables large language models (LLMs) to bypass the limitations of unidirectional causality, which underlies a phenomenon known as the reversal curse. When an LLM is prompted with sequential data, it often fails to recall preceding context. For example, when we ask an LLM to recall the line preceding “O say does that star-spangled banner yet wave” in the U.S. National Anthem, it often fails to correctly return “Gave proof through the night that our flag was still there”—this is due to the reversal curse. It occurs because language models such as ChatGPT and Llama generate text based on preceding tokens, requiring facts to be learned and reproduced in a consistent token order. While the reversal curse is often viewed as a limitation, we offer evidence of an alternative view: it is not always an obstacle in practice. We find that ReCall is driven by what we designate as cycle tokens—sequences that connect different parts of the training data, enabling recall of preceding tokens from succeeding ones. Through rigorous probabilistic formalization and controlled experiments, we demonstrate how the cycles they induce influence a model’s ability to reproduce information. To facilitate reproducibility, we provide our code and experimental details at https://anonymous.4open.science/r/remember-B0B8/.
%R 10.18653/v1/2025.acl-long.1232
%U https://aclanthology.org/2025.acl-long.1232/
%U https://doi.org/10.18653/v1/2025.acl-long.1232
%P 25365-25377
Markdown (Informal)
[Library-Like Behavior In Language Models is Enhanced by Self-Referencing Causal Cycles](https://aclanthology.org/2025.acl-long.1232/) (Nwadike et al., ACL 2025)
ACL
- Munachiso S Nwadike, Zangir Iklassov, Toluwani Aremu, Tatsuya Hiraoka, Benjamin Heinzerling, Velibor Bojkovic, Hilal AlQuabeh, Martin Takáč, and Kentaro Inui. 2025. Library-Like Behavior In Language Models is Enhanced by Self-Referencing Causal Cycles. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 25365–25377, Vienna, Austria. Association for Computational Linguistics.