@inproceedings{zhu-etal-2026-beyond-ranking,
title = "Beyond Ranking: Role-Specific Signals for Token-Budgeted {SFT} Data Selection",
author = "Zhu, Haocheng and
Wang, Tianqi and
Suzuki, Jun",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {EMNLP} 2026",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-emnlp.1956/",
doi = "10.18653/2026.findings-emnlp.1956",
pages = "41717--41729",
ISBN = "979-8-89176-462-0",
abstract = "Supervised fine-tuning (SFT) is a core stage in post-training general-purpose language models, but modern SFT training corpora are often large, heterogeneous mixtures, making data selection important under fixed token budgets. Simple ranking approaches can be informative yet insufficient: ranking alone does not determine how to allocate a fixed token budget across examples. We study the role-specific use of likelihood-derived signals for token-budgeted subset construction. An external instruction-tuned reference model computes response-side negative log-likelihood scores to rank training examples. We then combine this ranking with base model likelihood filtering to restrict the candidate region and response-side cosine filtering to reduce redundancy in the selected subset. In the OLMo fine-tuning setting studied here, ranking examples only by reference-model likelihood underperforms the random baseline, whereas the same ranking becomes more effective when combined with base model likelihood filtering and cosine-based redundancy control. Our results support viewing token-budgeted SFT data selection as a budget-allocation problem rather than only as an example-ranking problem."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="zhu-etal-2026-beyond-ranking">
<titleInfo>
<title>Beyond Ranking: Role-Specific Signals for Token-Budgeted SFT Data Selection</title>
</titleInfo>
<name type="personal">
<namePart type="given">Haocheng</namePart>
<namePart type="family">Zhu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tianqi</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jun</namePart>
<namePart type="family">Suzuki</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>Findings of the Association for Computational Linguistics: EMNLP 2026</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">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">979-8-89176-462-0</identifier>
</relatedItem>
<abstract>Supervised fine-tuning (SFT) is a core stage in post-training general-purpose language models, but modern SFT training corpora are often large, heterogeneous mixtures, making data selection important under fixed token budgets. Simple ranking approaches can be informative yet insufficient: ranking alone does not determine how to allocate a fixed token budget across examples. We study the role-specific use of likelihood-derived signals for token-budgeted subset construction. An external instruction-tuned reference model computes response-side negative log-likelihood scores to rank training examples. We then combine this ranking with base model likelihood filtering to restrict the candidate region and response-side cosine filtering to reduce redundancy in the selected subset. In the OLMo fine-tuning setting studied here, ranking examples only by reference-model likelihood underperforms the random baseline, whereas the same ranking becomes more effective when combined with base model likelihood filtering and cosine-based redundancy control. Our results support viewing token-budgeted SFT data selection as a budget-allocation problem rather than only as an example-ranking problem.</abstract>
<identifier type="citekey">zhu-etal-2026-beyond-ranking</identifier>
<identifier type="doi">10.18653/2026.findings-emnlp.1956</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.1956/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>41717</start>
<end>41729</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Beyond Ranking: Role-Specific Signals for Token-Budgeted SFT Data Selection
%A Zhu, Haocheng
%A Wang, Tianqi
%A Suzuki, Jun
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Findings of the Association for Computational Linguistics: EMNLP 2026
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-462-0
%F zhu-etal-2026-beyond-ranking
%X Supervised fine-tuning (SFT) is a core stage in post-training general-purpose language models, but modern SFT training corpora are often large, heterogeneous mixtures, making data selection important under fixed token budgets. Simple ranking approaches can be informative yet insufficient: ranking alone does not determine how to allocate a fixed token budget across examples. We study the role-specific use of likelihood-derived signals for token-budgeted subset construction. An external instruction-tuned reference model computes response-side negative log-likelihood scores to rank training examples. We then combine this ranking with base model likelihood filtering to restrict the candidate region and response-side cosine filtering to reduce redundancy in the selected subset. In the OLMo fine-tuning setting studied here, ranking examples only by reference-model likelihood underperforms the random baseline, whereas the same ranking becomes more effective when combined with base model likelihood filtering and cosine-based redundancy control. Our results support viewing token-budgeted SFT data selection as a budget-allocation problem rather than only as an example-ranking problem.
%R 10.18653/2026.findings-emnlp.1956
%U https://aclanthology.org/2026.findings-emnlp.1956/
%U https://doi.org/10.18653/2026.findings-emnlp.1956
%P 41717-41729
Markdown (Informal)
[Beyond Ranking: Role-Specific Signals for Token-Budgeted SFT Data Selection](https://aclanthology.org/2026.findings-emnlp.1956/) (Zhu et al., Findings 2026)
ACL