<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="asai-etal-2026-llms">
    <titleInfo>
        <title>Can LLMs Be Constrained to the Past? Improving Knowledge Cutoff through Recall-Based Prompting</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Michiro</namePart>
        <namePart type="family">Asai</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Ailiang</namePart>
        <namePart type="family">Lin</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Yu</namePart>
        <namePart type="family">Kishimoto</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Takao</namePart>
        <namePart type="family">Obi</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Satoshi</namePart>
        <namePart type="family">Kosugi</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Kotaro</namePart>
        <namePart type="family">Funakoshi</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Manabu</namePart>
        <namePart type="family">Okumura</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2026-10</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Hung-Yi</namePart>
            <namePart type="family">Lee</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Sunipa</namePart>
            <namePart type="family">Dev</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Junyi</namePart>
            <namePart type="given">Jessy</namePart>
            <namePart type="family">Li</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Desmond</namePart>
            <namePart type="family">Elliott</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Budapest, Hungary</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
        <identifier type="isbn">979-8-89176-459-0</identifier>
    </relatedItem>
    <abstract>Prompted knowledge cutoff instructs a large language model (LLM) to act as if information beyond a specified cutoff date were unavailable. However, prior work mainly relies on direct-answer generation, which struggles when post-cutoff knowledge is not explicitly queried but is only causally related to the question. To address this limitation, we propose two recall-based prompting strategies: Self-Recall (SR), which asks the model to restate its cutoff constraint, and Question-Recall (QR), which requires the model to recall question-relevant information valid under the cutoff. Across three existing benchmarks, our methods outperform both direct-answer prompting and conventional step-by-step reasoning baselines, with particularly strong improvements on counterfactual questions. To investigate robustness across different cutoff settings, we further construct the Multi-cutoff Historical Event Benchmark (MHEB), which evaluates the same question under multiple cutoff years. Results show that knowledge cutoff performance varies with cutoff distance, while combining SR and QR consistently yields the best performance. Code and dataset are available at [https://github.com/isct-lrlab/MHEB](https://github.com/isct-lrlab/MHEB).</abstract>
    <identifier type="citekey">asai-etal-2026-llms</identifier>
    <identifier type="doi">10.18653/2026.emnlp-main.895</identifier>
    <location>
        <url>https://aclanthology.org/2026.emnlp-main.895/</url>
    </location>
    <part>
        <date>2026-10</date>
        <extent unit="page">
            <start>20622</start>
            <end>20637</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Can LLMs Be Constrained to the Past? Improving Knowledge Cutoff through Recall-Based Prompting
%A Asai, Michiro
%A Lin, Ailiang
%A Kishimoto, Yu
%A Obi, Takao
%A Kosugi, Satoshi
%A Funakoshi, Kotaro
%A Okumura, Manabu
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Junyi Jessy
%Y Elliott, Desmond
%S Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-459-0
%F asai-etal-2026-llms
%X Prompted knowledge cutoff instructs a large language model (LLM) to act as if information beyond a specified cutoff date were unavailable. However, prior work mainly relies on direct-answer generation, which struggles when post-cutoff knowledge is not explicitly queried but is only causally related to the question. To address this limitation, we propose two recall-based prompting strategies: Self-Recall (SR), which asks the model to restate its cutoff constraint, and Question-Recall (QR), which requires the model to recall question-relevant information valid under the cutoff. Across three existing benchmarks, our methods outperform both direct-answer prompting and conventional step-by-step reasoning baselines, with particularly strong improvements on counterfactual questions. To investigate robustness across different cutoff settings, we further construct the Multi-cutoff Historical Event Benchmark (MHEB), which evaluates the same question under multiple cutoff years. Results show that knowledge cutoff performance varies with cutoff distance, while combining SR and QR consistently yields the best performance. Code and dataset are available at [https://github.com/isct-lrlab/MHEB](https://github.com/isct-lrlab/MHEB).
%R 10.18653/2026.emnlp-main.895
%U https://aclanthology.org/2026.emnlp-main.895/
%U https://doi.org/10.18653/2026.emnlp-main.895
%P 20622-20637

Markdown (Informal)

[Can LLMs Be Constrained to the Past? Improving Knowledge Cutoff through Recall-Based Prompting](https://aclanthology.org/2026.emnlp-main.895/) (Asai et al., EMNLP 2026)

• Can LLMs Be Constrained to the Past? Improving Knowledge Cutoff through Recall-Based Prompting (Asai et al., EMNLP 2026)

ACL

• Michiro Asai, Ailiang Lin, Yu Kishimoto, Takao Obi, Satoshi Kosugi, Kotaro Funakoshi, and Manabu Okumura. 2026. Can LLMs Be Constrained to the Past? Improving Knowledge Cutoff through Recall-Based Prompting. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, pages 20622–20637, Budapest, Hungary. Association for Computational Linguistics.