@inproceedings{ajlan-mohamed-2026-qahera,
title = "Qahera at {I}slamic{E}val 2026: Corpus-Grounded Retrieval and Iterative Harness Search with a 4{B}-Parameter {LLM} for Islamic Citation Verification",
author = "Ajlan, Hazem and
Mohamed, Mohamed",
editor = "Abdul-Mageed, Muhammad and
Elsayed, Tamer and
Bouamor, Houda and
El-Beltagy, Samhaa and
Al-Khalifa, Hend and
Abdelali, Ahmed and
Elmadany, AbdelRahim and
El Mekki, Abdellah and
Darwish, Kareem and
Alkhereyf, Sakhar and
Luqman, Hamzah and
Khalifa, Salam and
Magdy, Samar and
Alyafeai, Zaid and
Hamed, Injy and
Zaghouani, Wajdi and
Zitouni, Imed",
booktitle = "Proceedings of The Fourth {A}rabic Natural Language Processing Conference: Shared Tasks",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.arabicnlp-sharedtasks.134/",
pages = "1088--1092",
ISBN = "979-8-89176-497-2",
abstract = "We describe the Qahera system for the four IslamicEval 2026 subtasks: (1) detecting Qur{'}an/Hadith citation spans in LLM-generated Islamic answers, (2) judging citation authenticity, (3) correcting inauthentic citations against reference corpora, and (4) judging whether an authentic citation answers the question asked. Under a 13B-parameter open-model constraint, running entirely on free Kaggle GPUs, we combine a lightweight token-classification model with character-level auxiliary features (Sub-task 1), corpus-grounded fuzzy retrieval (Sub-tasks 2{--}3), and a 4B-parameter instruction-tuned LLM (Gemma-3-4B) used only where retrieval alone cannot discriminate (Subtasks 3{--}4). Rather than hand-designing LLM prompts once, we develop them through an iterative propose-evaluate-diagnose search loop that reads full prediction traces after every candidate and hunts for systematic biases before each change. This surfaces small-model failure modes {--} a near-universal ``reject'' bias in isolated yes/no verification, and an ``always name something'' bias in numbered-list prompts {--} that would otherwise silently cap performance, and shows that validating candidates on small subsets can be actively misleading. Our final system reaches 0.9716 F1 (Subtask 1), 0.906 accuracy (Subtask 2), 0.773 accuracy (Subtask 3), and 0.673 macro-F1 (Subtask 4) on the official test sets."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="ajlan-mohamed-2026-qahera">
<titleInfo>
<title>Qahera at IslamicEval 2026: Corpus-Grounded Retrieval and Iterative Harness Search with a 4B-Parameter LLM for Islamic Citation Verification</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hazem</namePart>
<namePart type="family">Ajlan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mohamed</namePart>
<namePart type="family">Mohamed</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 Fourth Arabic Natural Language Processing Conference: Shared Tasks</title>
</titleInfo>
<name type="personal">
<namePart type="given">Muhammad</namePart>
<namePart type="family">Abdul-Mageed</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tamer</namePart>
<namePart type="family">Elsayed</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Houda</namePart>
<namePart type="family">Bouamor</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Samhaa</namePart>
<namePart type="family">El-Beltagy</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hend</namePart>
<namePart type="family">Al-Khalifa</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ahmed</namePart>
<namePart type="family">Abdelali</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">AbdelRahim</namePart>
<namePart type="family">Elmadany</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Abdellah</namePart>
<namePart type="family">El Mekki</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kareem</namePart>
<namePart type="family">Darwish</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sakhar</namePart>
<namePart type="family">Alkhereyf</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hamzah</namePart>
<namePart type="family">Luqman</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Salam</namePart>
<namePart type="family">Khalifa</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Samar</namePart>
<namePart type="family">Magdy</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zaid</namePart>
<namePart type="family">Alyafeai</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Injy</namePart>
<namePart type="family">Hamed</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Wajdi</namePart>
<namePart type="family">Zaghouani</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Imed</namePart>
<namePart type="family">Zitouni</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">979-8-89176-497-2</identifier>
</relatedItem>
<abstract>We describe the Qahera system for the four IslamicEval 2026 subtasks: (1) detecting Qur’an/Hadith citation spans in LLM-generated Islamic answers, (2) judging citation authenticity, (3) correcting inauthentic citations against reference corpora, and (4) judging whether an authentic citation answers the question asked. Under a 13B-parameter open-model constraint, running entirely on free Kaggle GPUs, we combine a lightweight token-classification model with character-level auxiliary features (Sub-task 1), corpus-grounded fuzzy retrieval (Sub-tasks 2–3), and a 4B-parameter instruction-tuned LLM (Gemma-3-4B) used only where retrieval alone cannot discriminate (Subtasks 3–4). Rather than hand-designing LLM prompts once, we develop them through an iterative propose-evaluate-diagnose search loop that reads full prediction traces after every candidate and hunts for systematic biases before each change. This surfaces small-model failure modes – a near-universal “reject” bias in isolated yes/no verification, and an “always name something” bias in numbered-list prompts – that would otherwise silently cap performance, and shows that validating candidates on small subsets can be actively misleading. Our final system reaches 0.9716 F1 (Subtask 1), 0.906 accuracy (Subtask 2), 0.773 accuracy (Subtask 3), and 0.673 macro-F1 (Subtask 4) on the official test sets.</abstract>
<identifier type="citekey">ajlan-mohamed-2026-qahera</identifier>
<location>
<url>https://aclanthology.org/2026.arabicnlp-sharedtasks.134/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>1088</start>
<end>1092</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Qahera at IslamicEval 2026: Corpus-Grounded Retrieval and Iterative Harness Search with a 4B-Parameter LLM for Islamic Citation Verification
%A Ajlan, Hazem
%A Mohamed, Mohamed
%Y Abdul-Mageed, Muhammad
%Y Elsayed, Tamer
%Y Bouamor, Houda
%Y El-Beltagy, Samhaa
%Y Al-Khalifa, Hend
%Y Abdelali, Ahmed
%Y Elmadany, AbdelRahim
%Y El Mekki, Abdellah
%Y Darwish, Kareem
%Y Alkhereyf, Sakhar
%Y Luqman, Hamzah
%Y Khalifa, Salam
%Y Magdy, Samar
%Y Alyafeai, Zaid
%Y Hamed, Injy
%Y Zaghouani, Wajdi
%Y Zitouni, Imed
%S Proceedings of The Fourth Arabic Natural Language Processing Conference: Shared Tasks
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-497-2
%F ajlan-mohamed-2026-qahera
%X We describe the Qahera system for the four IslamicEval 2026 subtasks: (1) detecting Qur’an/Hadith citation spans in LLM-generated Islamic answers, (2) judging citation authenticity, (3) correcting inauthentic citations against reference corpora, and (4) judging whether an authentic citation answers the question asked. Under a 13B-parameter open-model constraint, running entirely on free Kaggle GPUs, we combine a lightweight token-classification model with character-level auxiliary features (Sub-task 1), corpus-grounded fuzzy retrieval (Sub-tasks 2–3), and a 4B-parameter instruction-tuned LLM (Gemma-3-4B) used only where retrieval alone cannot discriminate (Subtasks 3–4). Rather than hand-designing LLM prompts once, we develop them through an iterative propose-evaluate-diagnose search loop that reads full prediction traces after every candidate and hunts for systematic biases before each change. This surfaces small-model failure modes – a near-universal “reject” bias in isolated yes/no verification, and an “always name something” bias in numbered-list prompts – that would otherwise silently cap performance, and shows that validating candidates on small subsets can be actively misleading. Our final system reaches 0.9716 F1 (Subtask 1), 0.906 accuracy (Subtask 2), 0.773 accuracy (Subtask 3), and 0.673 macro-F1 (Subtask 4) on the official test sets.
%U https://aclanthology.org/2026.arabicnlp-sharedtasks.134/
%P 1088-1092
Markdown (Informal)
[Qahera at IslamicEval 2026: Corpus-Grounded Retrieval and Iterative Harness Search with a 4B-Parameter LLM for Islamic Citation Verification](https://aclanthology.org/2026.arabicnlp-sharedtasks.134/) (Ajlan & Mohamed, ArabicNLP 2026)
ACL