BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Los_Angeles
X-LIC-LOCATION:America/Los_Angeles
BEGIN:DAYLIGHT
TZOFFSETFROM:-0800
TZOFFSETTO:-0700
TZNAME:PDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0700
TZOFFSETTO:-0800
TZNAME:PST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260402T024534Z
LOCATION:Level 2 Lobby
DTSTART;TZID=America/Los_Angeles:20250623T180000
DTEND;TZID=America/Los_Angeles:20250623T190000
UID:dac_DAC 2025_sess287_LBR110@linklings.com
SUMMARY:Late Breaking Results: Fine-Tuning LLMs for Test Stimuli Generatio
 n
DESCRIPTION:Hyeonwoo Park, Seonghyeon Park, and Seokhyeong Kang (Pohang Un
 iversity of Science and Technology (POSTECH))\n\nThe understanding and rea
 soning capabilities of large language models (LLMs) with text data have ma
 de them widely used for test stimuli generation. Existing studies have pri
 marily focused on methods such as prompt engineering or providing feedback
  to the LLMs' generated outputs to improve test stimuli generation. Howeve
 r, these approaches have not been successful in enhancing the LLMs' domain
 -specific performance in generating test stimuli. In this paper, we introd
 uce a framework for fine- tuning LLMs for test stimuli generation through 
 dataset generation and reinforcement learning (RL). Our dataset generation
  approach creates a table-shaped test stimuli dataset, which helps ensure 
 that the LLM produces consistent outputs. Additionally, our two-stage fine
 -tuning process involves training the LLMs on domain-specific data and usi
 ng RL to provide feedback on the generated outputs, further enhancing the 
 LLMs' performance in test stimuli generation. Experimental results confirm
  that our framework improves syntax correctness and code coverage of test 
 stimuli, outperforming commercial models.\n\n
END:VEVENT
END:VCALENDAR