SSA: Improving Performance With a Better Scoring Function

Omar Naim; Swarnadeep Bhar; Jerome Bolte; Nicholas Asher

SSA: Improving Performance With a Better Scoring Function

Omar Naim, Swarnadeep Bhar, Jerome Bolte, Nicholas Asher

Correct Metadata for

Use this form to create a GitHub issue with structured data describing the correction. You will need a GitHub account. Once you create that issue, the correction will be reviewed by a staff member.

⚠️ Mobile Users: Submitting this form to create a new issue will only work with github.com, not the GitHub Mobile app.

Important: The Anthology treat PDFs as authoritative. Please use this form only to correct data that is out of line with the PDF. See our corrections guidelines if you need to change the PDF.

Title Adjust the title. Retain tags such as <fixed-case>.

Authors Adjust author names and order to match the PDF.

Abstract Correct abstract if needed. Retain XML formatting tags such as <tex-math>. You may use <b>...</b> for bold, <i>...</i> for italic, and <url>...</url> for URLs.

Verification against PDF Ensure that the new title/authors match the snapshot below. (If there is no snapshot or it is too small, consult the PDF.)

Authors concatenated from the text boxes above:

ALL author names match the snapshot above—including middle initials, hyphens, and accents.

Abstract

While transformer models exhibit strong in-context learning (ICL) abilities, they often fail to generalize under simple distribution shifts. We analyze these failures and identify Softmax, the scoring function in the attention mechanism, as a contributing factor. We propose Scaled Signed Averaging (SSA), a novel attention scoring function that mitigates these failures. SSA significantly improves performance on our ICL tasks and outperforms transformer models with Softmax on several NLP benchmarks and linguistic probing tasks, in both decoder-only and encoder-only architectures.

Anthology ID:: 2026.acl-long.1385
Volume:: Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
Month:: July
Year:: 2026
Address:: San Diego, California, United States
Editors:: Maria Liakata, Viviane P. Moreira, Jiajun Zhang, David Jurgens
Venue:: ACL
SIG:
Publisher:: Association for Computational Linguistics
Note:
Pages:: 30033–30051
Language:
URL:: https://aclanthology.org/2026.acl-long.1385/
DOI:
Bibkey:
Cite (ACL):: Omar Naim, Swarnadeep Bhar, Jerome Bolte, and Nicholas Asher. 2026. SSA: Improving Performance With a Better Scoring Function. In Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 30033–30051, San Diego, California, United States. Association for Computational Linguistics.
Cite (Informal):: SSA: Improving Performance With a Better Scoring Function (Naim et al., ACL 2026)
Copy Citation:
PDF:: https://aclanthology.org/2026.acl-long.1385.pdf
Checklist:: 2026.acl-long.1385.checklist.pdf

PDF Cite Search Checklist Fix data

Export citation

BibTeX
MODS XML
Endnote
Preformatted

@inproceedings{naim-etal-2026-ssa,
    title = "{SSA}: Improving Performance With a Better Scoring Function",
    author = "Naim, Omar  and
      Bhar, Swarnadeep  and
      Bolte, Jerome  and
      Asher, Nicholas",
    editor = "Liakata, Maria  and
      Moreira, Viviane P.  and
      Zhang, Jiajun  and
      Jurgens, David",
    booktitle = "Proceedings of the 64th Annual Meeting of the {A}ssociation for {C}omputational {L}inguistics (Volume 1: Long Papers)",
    month = jul,
    year = "2026",
    address = "San Diego, California, United States",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2026.acl-long.1385/",
    pages = "30033--30051",
    ISBN = "979-8-89176-390-6",
    abstract = "While transformer models exhibit strong in-context learning (ICL) abilities, they often fail to generalize under simple distribution shifts. We analyze these failures and identify Softmax, the scoring function in the attention mechanism, as a contributing factor. We propose Scaled Signed Averaging (SSA), a novel attention scoring function that mitigates these failures. SSA significantly improves performance on our ICL tasks and outperforms transformer models with Softmax on several NLP benchmarks and linguistic probing tasks, in both decoder-only and encoder-only architectures."
}

Download as File

<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="naim-etal-2026-ssa">
    <titleInfo>
        <title>SSA: Improving Performance With a Better Scoring Function</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Omar</namePart>
        <namePart type="family">Naim</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Swarnadeep</namePart>
        <namePart type="family">Bhar</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Jerome</namePart>
        <namePart type="family">Bolte</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Nicholas</namePart>
        <namePart type="family">Asher</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2026-07</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Maria</namePart>
            <namePart type="family">Liakata</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Viviane</namePart>
            <namePart type="given">P</namePart>
            <namePart type="family">Moreira</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Jiajun</namePart>
            <namePart type="family">Zhang</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">David</namePart>
            <namePart type="family">Jurgens</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">San Diego, California, United States</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
        <identifier type="isbn">979-8-89176-390-6</identifier>
    </relatedItem>
    <abstract>While transformer models exhibit strong in-context learning (ICL) abilities, they often fail to generalize under simple distribution shifts. We analyze these failures and identify Softmax, the scoring function in the attention mechanism, as a contributing factor. We propose Scaled Signed Averaging (SSA), a novel attention scoring function that mitigates these failures. SSA significantly improves performance on our ICL tasks and outperforms transformer models with Softmax on several NLP benchmarks and linguistic probing tasks, in both decoder-only and encoder-only architectures.</abstract>
    <identifier type="citekey">naim-etal-2026-ssa</identifier>
    <location>
        <url>https://aclanthology.org/2026.acl-long.1385/</url>
    </location>
    <part>
        <date>2026-07</date>
        <extent unit="page">
            <start>30033</start>
            <end>30051</end>
        </extent>
    </part>
</mods>
</modsCollection>

Download as File

%0 Conference Proceedings
%T SSA: Improving Performance With a Better Scoring Function
%A Naim, Omar
%A Bhar, Swarnadeep
%A Bolte, Jerome
%A Asher, Nicholas
%Y Liakata, Maria
%Y Moreira, Viviane P.
%Y Zhang, Jiajun
%Y Jurgens, David
%S Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
%D 2026
%8 July
%I Association for Computational Linguistics
%C San Diego, California, United States
%@ 979-8-89176-390-6
%F naim-etal-2026-ssa
%X While transformer models exhibit strong in-context learning (ICL) abilities, they often fail to generalize under simple distribution shifts. We analyze these failures and identify Softmax, the scoring function in the attention mechanism, as a contributing factor. We propose Scaled Signed Averaging (SSA), a novel attention scoring function that mitigates these failures. SSA significantly improves performance on our ICL tasks and outperforms transformer models with Softmax on several NLP benchmarks and linguistic probing tasks, in both decoder-only and encoder-only architectures.
%U https://aclanthology.org/2026.acl-long.1385/
%P 30033-30051

Download as File

Markdown (Informal)

[SSA: Improving Performance With a Better Scoring Function](https://aclanthology.org/2026.acl-long.1385/) (Naim et al., ACL 2026)

SSA: Improving Performance With a Better Scoring Function (Naim et al., ACL 2026)

ACL

Omar Naim, Swarnadeep Bhar, Jerome Bolte, and Nicholas Asher. 2026. SSA: Improving Performance With a Better Scoring Function. In Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 30033–30051, San Diego, California, United States. Association for Computational Linguistics.