@inproceedings{tamsal-2026-pfw,
title = "{PFW} at {S}em{E}val-2026 Task 6: Multi-Seed {D}e{BERT}a Ensembles for Political Response Clarity and Evasion Classification",
author = "Tamsal, Taleef",
editor = "Kochmar, Ekaterina and
Ghosh, Debanjan and
North, Kai and
Komachi, Mamoru",
booktitle = "Proceedings of the 20th {I}nternational {W}orkshop on {S}emantic {E}valuation (2026)",
month = jul,
year = "2026",
address = "San Diego, California, USA",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.semeval-1.197/",
pages = "1518--1525",
ISBN = "979-8-89176-414-9",
abstract = "This paper describes the PFW system for SemEval-2026 Task 6 (CLARITY), which addresses the classification of response clarity and evasion techniques in political interview question-answer pairs. Rather than relying on large language model prompting, we pursue a competitive non-LLM approach based on fine-tuning DeBERTa-xlarge and DeBERTa-v3-large with a multi-seed ensemble strategy: 5-fold cross-validation with 10 random seeds yields 50 models per architecture, combined through simple logit averaging. Our system achieves a macro F1 of 0.76 on Subtask 1 (clarity-level classification) and 0.50 on Subtask 2 (evasion-type classification). We also find that three post-hoc optimization techniques{---}learned ensemble weights, thresh old calibration, and hierarchical masking{---} each improve out-of-fold performance yet degrade evaluation scores by 0.02{--}0.10 F1. This pattern should be interpreted cautiously: the 237-sample evaluation set likely contributes substantial variance, and two of the three degradations fall within the {\ensuremath{\pm}}0.06 95{\%} CI expected from sampling noise. Still, the consistent directional pattern across all three prediction-level interventions provides suggestive evidence for an optimization paradox, highlighting the risk of overfitting to cross-validation predictions when evaluation data is limited. Our code is publicly available at https://github.com/ Taleef7/semeval-2026-task6."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="tamsal-2026-pfw">
<titleInfo>
<title>PFW at SemEval-2026 Task 6: Multi-Seed DeBERTa Ensembles for Political Response Clarity and Evasion Classification</title>
</titleInfo>
<name type="personal">
<namePart type="given">Taleef</namePart>
<namePart type="family">Tamsal</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 20th International Workshop on Semantic Evaluation (2026)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Ekaterina</namePart>
<namePart type="family">Kochmar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Debanjan</namePart>
<namePart type="family">Ghosh</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kai</namePart>
<namePart type="family">North</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mamoru</namePart>
<namePart type="family">Komachi</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, USA</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-414-9</identifier>
</relatedItem>
<abstract>This paper describes the PFW system for SemEval-2026 Task 6 (CLARITY), which addresses the classification of response clarity and evasion techniques in political interview question-answer pairs. Rather than relying on large language model prompting, we pursue a competitive non-LLM approach based on fine-tuning DeBERTa-xlarge and DeBERTa-v3-large with a multi-seed ensemble strategy: 5-fold cross-validation with 10 random seeds yields 50 models per architecture, combined through simple logit averaging. Our system achieves a macro F1 of 0.76 on Subtask 1 (clarity-level classification) and 0.50 on Subtask 2 (evasion-type classification). We also find that three post-hoc optimization techniques—learned ensemble weights, thresh old calibration, and hierarchical masking— each improve out-of-fold performance yet degrade evaluation scores by 0.02–0.10 F1. This pattern should be interpreted cautiously: the 237-sample evaluation set likely contributes substantial variance, and two of the three degradations fall within the \ensuremath\pm0.06 95% CI expected from sampling noise. Still, the consistent directional pattern across all three prediction-level interventions provides suggestive evidence for an optimization paradox, highlighting the risk of overfitting to cross-validation predictions when evaluation data is limited. Our code is publicly available at https://github.com/ Taleef7/semeval-2026-task6.</abstract>
<identifier type="citekey">tamsal-2026-pfw</identifier>
<location>
<url>https://aclanthology.org/2026.semeval-1.197/</url>
</location>
<part>
<date>2026-07</date>
<extent unit="page">
<start>1518</start>
<end>1525</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T PFW at SemEval-2026 Task 6: Multi-Seed DeBERTa Ensembles for Political Response Clarity and Evasion Classification
%A Tamsal, Taleef
%Y Kochmar, Ekaterina
%Y Ghosh, Debanjan
%Y North, Kai
%Y Komachi, Mamoru
%S Proceedings of the 20th International Workshop on Semantic Evaluation (2026)
%D 2026
%8 July
%I Association for Computational Linguistics
%C San Diego, California, USA
%@ 979-8-89176-414-9
%F tamsal-2026-pfw
%X This paper describes the PFW system for SemEval-2026 Task 6 (CLARITY), which addresses the classification of response clarity and evasion techniques in political interview question-answer pairs. Rather than relying on large language model prompting, we pursue a competitive non-LLM approach based on fine-tuning DeBERTa-xlarge and DeBERTa-v3-large with a multi-seed ensemble strategy: 5-fold cross-validation with 10 random seeds yields 50 models per architecture, combined through simple logit averaging. Our system achieves a macro F1 of 0.76 on Subtask 1 (clarity-level classification) and 0.50 on Subtask 2 (evasion-type classification). We also find that three post-hoc optimization techniques—learned ensemble weights, thresh old calibration, and hierarchical masking— each improve out-of-fold performance yet degrade evaluation scores by 0.02–0.10 F1. This pattern should be interpreted cautiously: the 237-sample evaluation set likely contributes substantial variance, and two of the three degradations fall within the \ensuremath\pm0.06 95% CI expected from sampling noise. Still, the consistent directional pattern across all three prediction-level interventions provides suggestive evidence for an optimization paradox, highlighting the risk of overfitting to cross-validation predictions when evaluation data is limited. Our code is publicly available at https://github.com/ Taleef7/semeval-2026-task6.
%U https://aclanthology.org/2026.semeval-1.197/
%P 1518-1525
Markdown (Informal)
[PFW at SemEval-2026 Task 6: Multi-Seed DeBERTa Ensembles for Political Response Clarity and Evasion Classification](https://aclanthology.org/2026.semeval-1.197/) (Tamsal, SemEval 2026)
ACL