@inproceedings{sarda-etal-2026-qraft,
title = "{QRAFT}: {QL}o{RA} Retrieval-Augmented Fine-Tuning for Causal Span Extraction in Financial Documents",
author = "Sarda, Bavya and
Chatwal, Pulkit and
Dabral, Sonal",
editor = "El-Haj, Mo and
Moreno Sandoval, Antonio and
Garcia-Serrano, Ana and
Chen, Chung-Chi and
Rayson, Paul and
Torterolo Orta, Yanco Amor and
Martinez, Paloma and
Porta, Jordi",
booktitle = "The 7th Financial Narrative Processing Workshop",
month = may,
year = "2026",
address = "Palma de Mallorca, Spain",
publisher = "European Language Resources Association (ELRA)",
url = "https://aclanthology.org/2026.fnp-1.19/",
doi = "10.63317/4a9pbop2c9nc",
pages = "175--180",
abstract = "Understanding why financial outcomes occur is as important as knowing what they are. Annual reports and regulatory filings are rich with causal reasoning, yet extracting that reasoning automatically remains a difficult problem {---} one that sits at the intersection of domain expertise, linguistic nuance, and machine comprehension. In this paper, we describe our participation in the English subtask of the Financial Document Causality Detection shared task, FinCausal 2026, where systems are asked to identify verbatim causal spans from financial paragraphs in response to abstractive causal questions. Our approach is grounded in the intuition that a small, well-adapted model with the right inductive biases can outperform a larger but unfocused one. We fine-tune Qwen2.5-4B-Instruct on 2,000 domain-annotated instances using QLoRA, a parameter-efficient technique that enables meaningful adaptation under modest computational resources. Before training, we reformat all instances into the Qwen ChatML instruction template to align the model{'}s generation behaviour with the verbatim extraction requirement of the task. At inference time, we further guide the model by retrieving the most causally relevant sentence from the context using TF-IDF cosine similarity, providing an explicit local signal before generation. Outputs are produced via greedy decoding to ensure deterministic, source-grounded predictions. Under the official LLM-as-a-judge evaluation framework {---} which scores responses on a 1{--}5 adequacy scale based on semantic correctness rather than lexical overlap {---} our system achieves a score of 4.76 out of 5, placing 4th out of nine teams on the English leaderboard. Our results suggest that combining instruction-tuned fine-tuning with lightweight retrieval is a practical and effective strategy for causal reasoning in specialised financial text."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="sarda-etal-2026-qraft">
<titleInfo>
<title>QRAFT: QLoRA Retrieval-Augmented Fine-Tuning for Causal Span Extraction in Financial Documents</title>
</titleInfo>
<name type="personal">
<namePart type="given">Bavya</namePart>
<namePart type="family">Sarda</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Pulkit</namePart>
<namePart type="family">Chatwal</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sonal</namePart>
<namePart type="family">Dabral</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-05</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>The 7th Financial Narrative Processing Workshop</title>
</titleInfo>
<name type="personal">
<namePart type="given">Mo</namePart>
<namePart type="family">El-Haj</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Antonio</namePart>
<namePart type="family">Moreno Sandoval</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ana</namePart>
<namePart type="family">Garcia-Serrano</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chung-Chi</namePart>
<namePart type="family">Chen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Paul</namePart>
<namePart type="family">Rayson</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yanco</namePart>
<namePart type="given">Amor</namePart>
<namePart type="family">Torterolo Orta</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Paloma</namePart>
<namePart type="family">Martinez</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jordi</namePart>
<namePart type="family">Porta</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>European Language Resources Association (ELRA)</publisher>
<place>
<placeTerm type="text">Palma de Mallorca, Spain</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Understanding why financial outcomes occur is as important as knowing what they are. Annual reports and regulatory filings are rich with causal reasoning, yet extracting that reasoning automatically remains a difficult problem — one that sits at the intersection of domain expertise, linguistic nuance, and machine comprehension. In this paper, we describe our participation in the English subtask of the Financial Document Causality Detection shared task, FinCausal 2026, where systems are asked to identify verbatim causal spans from financial paragraphs in response to abstractive causal questions. Our approach is grounded in the intuition that a small, well-adapted model with the right inductive biases can outperform a larger but unfocused one. We fine-tune Qwen2.5-4B-Instruct on 2,000 domain-annotated instances using QLoRA, a parameter-efficient technique that enables meaningful adaptation under modest computational resources. Before training, we reformat all instances into the Qwen ChatML instruction template to align the model’s generation behaviour with the verbatim extraction requirement of the task. At inference time, we further guide the model by retrieving the most causally relevant sentence from the context using TF-IDF cosine similarity, providing an explicit local signal before generation. Outputs are produced via greedy decoding to ensure deterministic, source-grounded predictions. Under the official LLM-as-a-judge evaluation framework — which scores responses on a 1–5 adequacy scale based on semantic correctness rather than lexical overlap — our system achieves a score of 4.76 out of 5, placing 4th out of nine teams on the English leaderboard. Our results suggest that combining instruction-tuned fine-tuning with lightweight retrieval is a practical and effective strategy for causal reasoning in specialised financial text.</abstract>
<identifier type="citekey">sarda-etal-2026-qraft</identifier>
<identifier type="doi">10.63317/4a9pbop2c9nc</identifier>
<location>
<url>https://aclanthology.org/2026.fnp-1.19/</url>
</location>
<part>
<date>2026-05</date>
<extent unit="page">
<start>175</start>
<end>180</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T QRAFT: QLoRA Retrieval-Augmented Fine-Tuning for Causal Span Extraction in Financial Documents
%A Sarda, Bavya
%A Chatwal, Pulkit
%A Dabral, Sonal
%Y El-Haj, Mo
%Y Moreno Sandoval, Antonio
%Y Garcia-Serrano, Ana
%Y Chen, Chung-Chi
%Y Rayson, Paul
%Y Torterolo Orta, Yanco Amor
%Y Martinez, Paloma
%Y Porta, Jordi
%S The 7th Financial Narrative Processing Workshop
%D 2026
%8 May
%I European Language Resources Association (ELRA)
%C Palma de Mallorca, Spain
%F sarda-etal-2026-qraft
%X Understanding why financial outcomes occur is as important as knowing what they are. Annual reports and regulatory filings are rich with causal reasoning, yet extracting that reasoning automatically remains a difficult problem — one that sits at the intersection of domain expertise, linguistic nuance, and machine comprehension. In this paper, we describe our participation in the English subtask of the Financial Document Causality Detection shared task, FinCausal 2026, where systems are asked to identify verbatim causal spans from financial paragraphs in response to abstractive causal questions. Our approach is grounded in the intuition that a small, well-adapted model with the right inductive biases can outperform a larger but unfocused one. We fine-tune Qwen2.5-4B-Instruct on 2,000 domain-annotated instances using QLoRA, a parameter-efficient technique that enables meaningful adaptation under modest computational resources. Before training, we reformat all instances into the Qwen ChatML instruction template to align the model’s generation behaviour with the verbatim extraction requirement of the task. At inference time, we further guide the model by retrieving the most causally relevant sentence from the context using TF-IDF cosine similarity, providing an explicit local signal before generation. Outputs are produced via greedy decoding to ensure deterministic, source-grounded predictions. Under the official LLM-as-a-judge evaluation framework — which scores responses on a 1–5 adequacy scale based on semantic correctness rather than lexical overlap — our system achieves a score of 4.76 out of 5, placing 4th out of nine teams on the English leaderboard. Our results suggest that combining instruction-tuned fine-tuning with lightweight retrieval is a practical and effective strategy for causal reasoning in specialised financial text.
%R 10.63317/4a9pbop2c9nc
%U https://aclanthology.org/2026.fnp-1.19/
%U https://doi.org/10.63317/4a9pbop2c9nc
%P 175-180
Markdown (Informal)
[QRAFT: QLoRA Retrieval-Augmented Fine-Tuning for Causal Span Extraction in Financial Documents](https://aclanthology.org/2026.fnp-1.19/) (Sarda et al., FNP 2026)
ACL