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:20260730T152640Z
LOCATION:
DTSTART;TZID=America/Los_Angeles:20260728T173400
DTEND;TZID=America/Los_Angeles:20260728T173500
UID:dac_DAC 2026_sess317_RESEARCH1042@linklings.com
SUMMARY:GearLS: Generalizable Reinforcement Learning Framework for Logic O
 ptimization via Policy Similarity Metric
DESCRIPTION:Yusen Qin, Jiaqi Lyu, Zhi Li, and Peng Cao (Southeast Universi
 ty)\n\nLogic optimization is a crucial step in digital circuit synthesis, 
 directly impacting the final power, performance, and area (PPA) of integra
 ted circuits. While reinforcement learning (RL) has shown promise in gener
 ating high-quality optimization flows, its limited generalization ability 
 necessitates retraining for each new circuit, hindering practical deployme
 nt. To address this, we propose a novel RL-based framework that achieves s
 trong zero-shot generalization across unseen circuits. Our work introduces
  three key innovations: (1) logic cone extraction to reduce input complexi
 ty and enable efficient reward estimation; (2) a cut-weighting mechanism t
 hat models global timing effects from local subgraphs; and (3) the integra
 tion of Policy Similarity Metric (PSM) to enhance state representation and
  improve zero-shot transfer. Evaluated on a set of unseen benchmark circui
 ts, Our work outperforms state-of-the-art methods—achieving 31.7% lower wo
 rst negative slack (WNS) and 32.5% lower total negative slack (TNS)—while 
 running in only 13% of the time required by prior approaches. This work de
 monstrates that generalizable RL can enable fast, high-quality logic optim
 ization without circuit-specific retraining.\n\nTrack: EDA7-II. Physical D
 esign and Verification, Design\n\nSession Chairs: Christian Pilato (Polite
 cnico di Milano); Mohsen Imani (University of California, Irvine); Flavio 
 Ponzina (San Diego State University); Amin Farshidi (Cadence Design System
 s, Inc.); Yonggan Fu (Georgia Institute of Technology, NVIDIA); Dimitrios 
 Tsaras (Huawei, Hong Kong University of Science and Technology (HKUST)); A
 nh Nguyen (Fujitsu Limited); Ruiyang Qin (Villanova University); Jiliang Z
 hang (Hunan University); Meng Li (Peking University); and Rassul Bairamkul
 ov (Advanced Micro Devices (AMD))\n\n
END:VEVENT
END:VCALENDAR