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:20260402T024533Z
LOCATION:Level 3 Lobby
DTSTART;TZID=America/Los_Angeles:20250622T180000
DTEND;TZID=America/Los_Angeles:20250622T190000
UID:dac_DAC 2025_sess261_RESEARCH2453@linklings.com
SUMMARY:OpenAssert: Towards Open-Source Large Language Models for Assertio
 n Generation
DESCRIPTION:Anand Menon, Samit Miftah, and Amisha Srivastava (The Universi
 ty of Texas at Dallas); Shamik Kundu, Arnab Raha, Souvik Kundu, Suvadeep B
 anerjee, and Deepak Mathaikutty (Intel Corporation); and Kanad Basu (The U
 niversity of Texas at Dallas)\n\nAssertions are essential for hardware ver
 ification but are typically generated manually, leading to long developmen
 t cycles. While commercial Large Language Models (LLMs) like GPT-4 show pr
 omise for automating assertion generation, they raise concerns about IP pr
 ivacy and data confidentiality. This paper proposes OpenAssert, an approac
 h for generating assertions locally using open-source LLMs. We enhance the
 se models with Retrieval Augmentation Generation (RAG) to reduce errors an
 d hallucinations. OpenAssert improves by up to 44% in rouge-1 score, 49% i
 n cosine similarity, and reduces word error rate by 43.4%, outperforming G
 PT-4 by 23.7%, with 100% line coverage in our evaluation.\n\nTracks: DES5:
  Emerging Device and Interconnect Technologies\n\n
END:VEVENT
END:VCALENDAR