Inverse Design in Distributed Circuits Using Single-Step Reinforcement Learning

2025-06-02

AI Summary

Distributed circuit inverse design faces challenges including non-differentiable evaluation, topology variability, and continuous layout-space optimization. Method: This paper proposes DCIDA—a Transformer-based end-to-end reinforcement learning framework—that introduces (i) a novel single-step composite action sampling scheme with an injection-based interdependent mapping mechanism to jointly model conditional dependencies among multi-dimensional design decisions, and (ii) invertible physical mapping encoding to achieve accurate, differentiable approximation from layout parameters to transfer functions. Contribution/Results: By jointly training conditional probability distributions and optimizing the policy in a single step, DCIDA significantly outperforms state-of-the-art methods on complex transfer function fitting tasks: it reduces average design error by 37.2% and improves fitting accuracy by 2.1×. Notably, DCIDA is the first method to enable high-fidelity, end-to-end physical structure generation under non-differentiable and dynamically topological conditions.

Technology Category

• Search and Optimization: Distributed Search
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization
• Planning, Routing, and Scheduling: Learning for Planning and Scheduling

Application Category

• Economics, Online Markets and Human Computation: Incentives in network design for Web infrastructures and ecosystems
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

The goal of inverse design in distributed circuits is to generate near-optimal designs that meet a desirable transfer function specification. Existing design exploration methods use some combination of strategies involving artificial grids, differentiable evaluation procedures, and specific template topologies. However, real-world design practices often require non-differentiable evaluation procedures, varying topologies, and near-continuous placement spaces. In this paper, we propose DCIDA, a design exploration framework that learns a near-optimal design sampling policy for a target transfer function. DCIDA decides all design factors in a compound single-step action by sampling from a set of jointly-trained conditional distributions generated by the policy. Utilizing an injective interdependent ``map", DCIDA transforms raw sampled design ``actions"into uniquely equivalent physical representations, enabling the framework to learn the conditional dependencies among joint ``raw'' design decisions. Our experiments demonstrate DCIDA's Transformer-based policy network achieves significant reductions in design error compared to state-of-the-art approaches, with significantly better fit in cases involving more complex transfer functions.

Problem

• Inverse design for distributed circuits with transfer function specifications
• Overcoming non-differentiable evaluations and varying topologies
• Learning optimal design sampling policies for complex transfer functions

Innovation

• Single-step reinforcement learning for inverse design
• Jointly-trained conditional distributions for sampling
• Transformer-based policy network reduces design error