# HeaRT: A Hierarchical Circuit Reasoning Tree-Based Agentic Framework for AMS Design Optimization

Source: [HeaRT: A Hierarchical Circuit Reasoning Tree-Based Agentic Framework for AMS Design Optimization](<https://www.sota2.com/research/paper/692689ead6b0dd33429306c3>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-03-31T05:00:04.048Z

Authors: Souradip Poddar, Chia-Tung Ho, Ziming Wei, Weidong Cao, Haoxing Ren, David Z. Pan

Published: 2025-11-24T20:11:06.000Z

## Abstract

Conventional AI-driven AMS design automation algorithms remain constrained by their reliance on high-quality datasets to capture underlying circuit behavior, coupled with poor transferability across architectures, and a lack of adaptive mechanisms. This work proposes HeaRT, a hierarchical circuit reasoning-based agentic framework for automation loops and a step toward adaptive, human-style design optimization. HeaRT consistently improves F1(subcircuits) by \>= 13.5% and F1(loops) by \>= 37.8% over few-shot prompting baselines across multiple LLM backbones on our 40-circuit AMS benchmark of flattened SPICE netlists, even as circuit complexity increases. Our experiments further show that HeaRT achieves \>= 3x faster convergence in incremental design adaptation tasks under specification shifts across diverse optimization approaches, supporting both topology reconfiguration and sizing.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Feedback Loop Identification on AMS benchmark Simple (\<20 transistors) 1.0](<https://www.sota2.com/research/sota/feedback-loop-identification-on-ams-benchmark-simple-20-transistors-1-0.md>): Feedback Loop Identification; AMS benchmark Simple (\<20 transistors) 1.0; F1 Score: 98.4.

- [Feedback Loop Identification on AMS benchmark Medium (20–40 transistors) 1.0](<https://www.sota2.com/research/sota/feedback-loop-identification-on-ams-benchmark-medium-20-40-transistors-1-0.md>): Feedback Loop Identification; AMS benchmark Medium (20–40 transistors) 1.0; F1 Score: 96.2.

- [Feedback Loop Identification on AMS 1.0 (Hard (\>40 transistors))](<https://www.sota2.com/research/sota/feedback-loop-identification-on-ams-1-0-hard-40-transistors.md>): Feedback Loop Identification; AMS 1.0 (Hard (\>40 transistors)); F1 Score: 93.7.

- [Subcircuit Identification on AMS benchmark Simple (\<20 transistors) 1.0](<https://www.sota2.com/research/sota/subcircuit-identification-on-ams-benchmark-simple-20-transistors-1-0.md>): Subcircuit Identification; AMS benchmark Simple (\<20 transistors) 1.0; F1 Score: 99.7.

- [Subcircuit Identification on AMS benchmark Medium (20–40 transistors) 1.0](<https://www.sota2.com/research/sota/subcircuit-identification-on-ams-benchmark-medium-20-40-transistors-1-0.md>): Subcircuit Identification; AMS benchmark Medium (20–40 transistors) 1.0; F1 Score: 98.8.

- [Subcircuit Identification on AMS Hard (\>40 transistors) 1.0](<https://www.sota2.com/research/sota/subcircuit-identification-on-ams-hard-40-transistors-1-0.md>): Subcircuit Identification; AMS Hard (\>40 transistors) 1.0; F1 Score: 95.5.

- [Analog circuit sizing optimization on Curated dataset repository Scenario 1: Relaxation Oscillator](<https://www.sota2.com/research/sota/analog-circuit-sizing-optimization-on-curated-dataset-repository-scenario-1-relaxation-oscillator.md>): Analog circuit sizing optimization; Curated dataset repository Scenario 1: Relaxation Oscillator; FoM: 618.97.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2511.19669>)

[PDF](<https://arxiv.org/pdf/2511.19669.pdf>)
