LLM Code Customization with Visual Results: A Benchmark on TikZ (EASE 2025 - AI Models / Data)

Tue 17 - Fri 20 June 2025 Istanbul, Turkey

Who

Charly Reux, Mathieu Acher, Djamel Eddine Khelladi, Clément Quinton, Olivier Barais

Track

EASE 2025 AI Models / Data

Time Zone

The program is currently displayed in (GMT+03:00) Athens.

Use conference time zone: (GMT+03:00) AthensSelect other time zone

The GMT offsets shown reflect the offsets at the moment of the conference.

Time Band

By setting a time band, the program will dim events that are outside this time window. This is useful for (virtual) conferences with a continuous program (with repeated sessions).
The time band will also limit the events that are included in the personal iCalendar subscription service.

Display full programSpecify a time band

Save

When

Wed 18 Jun 2025 14:45 - 15:00 at Senate Hall - LLMs for SE (Code) Chair(s): Giuseppe Destefanis

Link to Publication

https://arxiv.org/abs/2505.04670

Link to Preprint

https://arxiv.org/abs/2505.04670

File attachments

Presentation (EASE - LLM code customization TikZ Benchmark (1).pdf)	2.20MiB

Charly Reux

Univ Rennes, Inria, IRISA, INSA

France

Mathieu Acher

Univ Rennes, Inria, CNRS, IRISA

France

Djamel Eddine Khelladi

CNRS, IRISA, University of Rennes

France

Clément Quinton

University of Lille

France

Olivier Barais

University of Rennes; Inria; CNRS; IRISA

France

Time Zone

The program is currently displayed in (GMT+03:00) Athens.

Use conference time zone: (GMT+03:00) AthensSelect other time zone

The GMT offsets shown reflect the offsets at the moment of the conference.

Time Band

Display full programSpecify a time band

Save

Session Program

Wed 18 Jun
Displayed time zone: Athens change

13:30 - 15:00	LLMs for SE (Code)Research Papers / AI Models / Data at Senate Hall Chair(s): Giuseppe Destefanis Brunel University of London

13:30 15m Talk		CallNavi, A challenge and empirical study on LLM function calling and routing Research Papers Yewei Song University of Luxembourg, Xunzhu Tang University of Luxembourg, Cedric Lothritz Luxembourg Institute of Science and Technology, Saad Ezzini King Fahad University of Petroleum and Minerals, Jacques Klein University of Luxembourg, Tegawendé F. Bissyandé University of Luxembourg, Andrey Boytsov BGL BNP PARIBAS, Ulrick Ble BGL BNP PARIBAS, Anne Goujon BGL BNP PARIBAS Pre-print
13:45 15m Talk		CCCI: Code Completion with Contextual Information for Complex Data Transfer Tasks Using Large Language Models Research Papers Hangzhan Jin Polytechnique Montreal, Mohammad Hamdaqa Polytechnique Montreal
14:00 15m Research paper		Do Automatic Comment Generation Techniques Fall Short? Exploring the Influence of Method Dependencies on Code Understanding Research Papers Md Mustakim Billah University of Saskatchewan, Md Shamimur Rahman University of Saskatchewan, Banani Roy University of Saskatchewan Pre-print
14:15 15m Talk		How Well Do Large Language Models Serve as End-to-End Secure Code Agents for Python? AI Models / Data Jianian Gong Beihang University, Nachuan Duan Beihang University, Ziheng Tao Beihang University, Zhaohui Gong Beihang University, Yuan Yuan Beihang University, Minlie Huang Tsinghua University
14:30 15m Talk		Peft: Multiline complex patch correctness assessment based on fine-tuning Large Language Model with "Gold Data" AI Models / Data Xiaoxi Zheng Beijing University of Chemical Technology, Ruilian Zhao Beijing University of Chemical Technology, Junxia Guo Beijing University of Chemical Technology
14:45 15m Talk		LLM Code Customization with Visual Results: A Benchmark on TikZ AI Models / Data Charly Reux Univ Rennes, Inria, IRISA, INSA, Mathieu Acher Univ Rennes, Inria, CNRS, IRISA, Djamel Eddine Khelladi CNRS, IRISA, University of Rennes, Clément Quinton University of Lille, Olivier Barais University of Rennes; Inria; CNRS; IRISA Link to publication Pre-print File Attached