서울대학교 AI연구원과 Simular가 공동 주최하는 ICML 2026 주간 특별 연구 포럼에 여러분을 초대합니다.

Computer-Use Agents(CUA)의 최신 연구, 제품 동향, 실제 활용 사례를 한 자리에서 만나볼 수 있는 행사로, Simular 연구진과 서울대학교 연구진이 함께 차세대 AI 에이전트의 발전 방향을 논의합니다.
신뢰성과 안정성을 갖춘 AI 에이전트 구축 방법, Agent S3 연구 성과, 실제 업무 환경 적용 사례 등을 소개하며, 패널 토론과 Q&A 세션도 마련되어 있습니다.

📍 서울대학교 301동 118호
🗣 영어 진행
🎁 다과 및 참가자 기념품 제공
⚠ 좌석 한정, 사전 등록 필수

​A research-led fringe event during ICML 2026 week, co-hosted by Seoul National University AI Institute (SNU AIIS) and Simular.

​Four short talks spanning product, research, and field practice on computer-use agents (CUAs), followed by a fireside and open Q&A.
​Trust, control, and reliability are treated as first-class technical problems — verifiable outputs, on-device execution, guardrails, and audit trails.

REGISTRATION
https://luma.com/simular-l4tr

​WHO IT'S FOR
​SNU students, graduate researchers, and builders working on or curious about the frontier of computer-use agents. Seating is limited.

​ON THE DAY
​Light bites provided. Merch for attendees.

​WHAT YOU'LL TAKE AWAY
​• A clear mental model of where CUAs are today
​• The research arc behind reliable agents — Agent S → S2 → S3
​• How trustworthiness is engineered into computer-use
​• Concrete patterns for taking agents from research into real workflows

​RUN OF SHOW (9:00–11:00 KST)
​• 08:45 — Registration opens
​• 09:00 — Welcome & academic framing (Prof. Jonghyun Choi, SNU)
​• 09:10 — Product & landscape (Ang Li, Simular)
​• 09:25 — Research: Agent S3 (Eric Wang, Simular)
​• 09:40 — Trustworthiness in CUAs (Qingni Wang, Simular)
​• 09:50 — From research to real workflows (Robin Wu, Simular)
​• 10:00 — Buffer / stage reset
​• 10:05 — Fireside
​• 10:50 — Open Q&A

​FIRESIDE PANEL
​Moderated by Rita Liao (Simular), with Prof. Jonghyun Choi (SNU), Ang Li (Simular), and Dongkeun Jo (JoCoding; Korean AI-coding YouTuber).

​VENUE
Room 118, Building 301, Seoul National Universit

​Language
English