@inproceedings{ulli-etal-2026-aurum,
title = "Aurum at {CRF} Filling 2026: Modular {DSP}y Extractors with Qwen3-Max for Multilingual {CRF} Filling",
author = "Ulli, Vinay Babu and
Kumari, Jyoti and
Mondal, Anindita",
editor = "Gupta, Deepak and
Thompson, Paul and
Ananiadou, Sophia and
Demner-Fushman, Dina",
booktitle = "Proceedings of the Third Workshop on Patient-Oriented Language Processing ({CL}4{H}ealth) @ {LREC} 2026",
month = may,
year = "2026",
address = "Palma, Mallorca (Spain)",
publisher = "ELRA Language Resources Association (ELRA)",
url = "https://aclanthology.org/2026.cl4health-1.36/",
doi = "10.63317/4h5ugiqk5eht",
pages = "395--401",
abstract = "This paper describes the submission by Team Aurum to the CL4Health @ LREC 2026 Shared Task on Case Report Form (CRF) Filling from dyspnea patient clinical notes. Extracting 134 structured clinical fields using a single Large Language Model (LLM) call often leads to schema-following errors, hallucination, and poor attention over complex instructions. To address this, we propose a modular extraction pipeline built with DSPy, which decomposes the 134 CRF fields into 14 specialized, domain-specific extractors (e.g., Medical History, Lab Values, Acute Diagnoses). We conducted extensive experiments across multiple multilingual LLMs, including Llama4 Maverik, GPT-4o, GPT-4o Mini, DeepSeek-V3, Gemma-3-12B-Instruct, and Qwen-series models. Among these, Qwen3-Max (Thinking) with our optimized v2 prompts achieved the best performance on the development set with a Macro-F1 of 0.70, outperforming other evaluated models such as GPT-4o (0.68) and DeepSeek-V3 (0.66). Prompt optimization resulted in measurable gains, improving Qwen3-Max performance from 0.67 to 0.70. Using this configuration, our pipeline achieved an official Codabench Test Macro-F1 score of 0.68 in English and 0.67 in Italian, securing the 1st place ranking overall in the shared task."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="ulli-etal-2026-aurum">
<titleInfo>
<title>Aurum at CRF Filling 2026: Modular DSPy Extractors with Qwen3-Max for Multilingual CRF Filling</title>
</titleInfo>
<name type="personal">
<namePart type="given">Vinay</namePart>
<namePart type="given">Babu</namePart>
<namePart type="family">Ulli</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jyoti</namePart>
<namePart type="family">Kumari</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Anindita</namePart>
<namePart type="family">Mondal</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>Proceedings of the Third Workshop on Patient-Oriented Language Processing (CL4Health) @ LREC 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Deepak</namePart>
<namePart type="family">Gupta</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Paul</namePart>
<namePart type="family">Thompson</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sophia</namePart>
<namePart type="family">Ananiadou</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Dina</namePart>
<namePart type="family">Demner-Fushman</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>ELRA Language Resources Association (ELRA)</publisher>
<place>
<placeTerm type="text">Palma, Mallorca (Spain)</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>This paper describes the submission by Team Aurum to the CL4Health @ LREC 2026 Shared Task on Case Report Form (CRF) Filling from dyspnea patient clinical notes. Extracting 134 structured clinical fields using a single Large Language Model (LLM) call often leads to schema-following errors, hallucination, and poor attention over complex instructions. To address this, we propose a modular extraction pipeline built with DSPy, which decomposes the 134 CRF fields into 14 specialized, domain-specific extractors (e.g., Medical History, Lab Values, Acute Diagnoses). We conducted extensive experiments across multiple multilingual LLMs, including Llama4 Maverik, GPT-4o, GPT-4o Mini, DeepSeek-V3, Gemma-3-12B-Instruct, and Qwen-series models. Among these, Qwen3-Max (Thinking) with our optimized v2 prompts achieved the best performance on the development set with a Macro-F1 of 0.70, outperforming other evaluated models such as GPT-4o (0.68) and DeepSeek-V3 (0.66). Prompt optimization resulted in measurable gains, improving Qwen3-Max performance from 0.67 to 0.70. Using this configuration, our pipeline achieved an official Codabench Test Macro-F1 score of 0.68 in English and 0.67 in Italian, securing the 1st place ranking overall in the shared task.</abstract>
<identifier type="citekey">ulli-etal-2026-aurum</identifier>
<identifier type="doi">10.63317/4h5ugiqk5eht</identifier>
<location>
<url>https://aclanthology.org/2026.cl4health-1.36/</url>
</location>
<part>
<date>2026-05</date>
<extent unit="page">
<start>395</start>
<end>401</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Aurum at CRF Filling 2026: Modular DSPy Extractors with Qwen3-Max for Multilingual CRF Filling
%A Ulli, Vinay Babu
%A Kumari, Jyoti
%A Mondal, Anindita
%Y Gupta, Deepak
%Y Thompson, Paul
%Y Ananiadou, Sophia
%Y Demner-Fushman, Dina
%S Proceedings of the Third Workshop on Patient-Oriented Language Processing (CL4Health) @ LREC 2026
%D 2026
%8 May
%I ELRA Language Resources Association (ELRA)
%C Palma, Mallorca (Spain)
%F ulli-etal-2026-aurum
%X This paper describes the submission by Team Aurum to the CL4Health @ LREC 2026 Shared Task on Case Report Form (CRF) Filling from dyspnea patient clinical notes. Extracting 134 structured clinical fields using a single Large Language Model (LLM) call often leads to schema-following errors, hallucination, and poor attention over complex instructions. To address this, we propose a modular extraction pipeline built with DSPy, which decomposes the 134 CRF fields into 14 specialized, domain-specific extractors (e.g., Medical History, Lab Values, Acute Diagnoses). We conducted extensive experiments across multiple multilingual LLMs, including Llama4 Maverik, GPT-4o, GPT-4o Mini, DeepSeek-V3, Gemma-3-12B-Instruct, and Qwen-series models. Among these, Qwen3-Max (Thinking) with our optimized v2 prompts achieved the best performance on the development set with a Macro-F1 of 0.70, outperforming other evaluated models such as GPT-4o (0.68) and DeepSeek-V3 (0.66). Prompt optimization resulted in measurable gains, improving Qwen3-Max performance from 0.67 to 0.70. Using this configuration, our pipeline achieved an official Codabench Test Macro-F1 score of 0.68 in English and 0.67 in Italian, securing the 1st place ranking overall in the shared task.
%R 10.63317/4h5ugiqk5eht
%U https://aclanthology.org/2026.cl4health-1.36/
%U https://doi.org/10.63317/4h5ugiqk5eht
%P 395-401
Markdown (Informal)
[Aurum at CRF Filling 2026: Modular DSPy Extractors with Qwen3-Max for Multilingual CRF Filling](https://aclanthology.org/2026.cl4health-1.36/) (Ulli et al., CL4Health 2026)
ACL