Pith. sign in

REVIEW 4 major objections 5 minor 1 cited by

PPAAS: PVT and Pareto Aware Analog Sizing via Goal-conditioned Reinforcement Learning

T0 review · 4 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read A goal-conditioned reinforcement learning agent, trained with Pareto-front goal sampling and conservative hindsight replay, achieves roughly 1.6x sample efficiency and 4.1x simulation efficiency for PVT-aware analog sizing.

desk verdict A solid, reproducible GCRL recipe for PVT-aware analog sizing with honest ablations; the headline ratios rest on some slightly ad hoc baseline tweaks, and the main stress-test worry about Eq. (3) does not hold up. read the letter →

arxiv 2507.17003 v2 pith:SYZKRWF6 submitted 2025-07-22 eess.SP

Seunggeun Kim , Ziyi Wang , Sungyoung Lee , Youngmin Oh , Hanqing Zhu , Doyun Kim , David Z. Pan This is my paper · ORCID
classification eess.SP
keywords analogcircuitsizingPVTvariationsgoal-conditionedreinforcementlearningParetodominancegoalsamplinghindsightexperiencereplaysoftactor-criticsampleefficiencysimulation
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper tries to show that analog device sizing under process, voltage, and temperature (PVT) variations can be made substantially cheaper to train for without losing generalization to new design targets. It proposes PPAAS, a goal-conditioned reinforcement learning framework in which one policy is trained over a space of performance goals rather than retrained per specification. The framework's three main pieces are a Skip-on-Fail simulation schedule that runs full PVT corner simulations only when the nominal corner already meets the goal, a Pareto-front Dominance Goal Sampling scheme that draws training goals from the frontier of previously achieved goals, and a Conservative Hindsight Experience Replay that relabels failed episodes with conservative rewards. On four benchmark circuits, the paper reports success rates of 92.6%, 89.3%, 88.4%, and 69.6%, with roughly 1.6x sample efficiency and 4.1x simulation efficiency over prior sizing methods.

What carries the argument

The load-bearing objects are three interacting mechanisms. Skip-on-Fail (SoF) is a hierarchical simulation gate: stage 1 simulates only the nominal corner; if the nominal metrics satisfy the target, stage 2 simulates all other PVT corners and records a column-wise worst-case vector Worst(Zt) plus two binary success flags; if the nominal corner fails, the full-corner simulation is skipped. Pareto-front Dominance Goal Sampling (PGDS) maintains a buffer of previously achieved goals, samples candidates that are not Pareto-dominated by that buffer, and selects among them by softmax over negative Q-estimates, so the agent is pushed toward goals that are hard but reachable. Conservative Hindsight Experience Replay (CHER) adds a relabeled transition for each real transition, using a future achieved goal and a deliberately lower reward scale R' ≤ R so that the virtual signal never overstates how good the relabeled goal is. The reward is a three-stage hierarchical function tied to the SoF stage, with an optional PVT-consistency penalty that penalizes squared relative deviations from nominal-corner metrics.

What would settle it

Evaluate the trained policy on a held-out set of goals in which the nominal corner is the only failing corner while all extreme corners meet the specification; because Skip-on-Fail classifies these transitions as nominal-stage failures and records Dt = 0, a policy that still succeeds on such goals would contradict the claim that the training signal preserves corner-level success information, while a systematic failure would confirm the supervision is incomplete.

Watch

Extended reading notes

Core claim

The central claim is that the combination of Pareto-front goal sampling, conservative hindsight relabeling, and a nominal-first simulation gate makes a goal-conditioned Soft Actor-Critic policy learn PVT-robust sizing from fewer samples and fewer simulations than existing GCRL or single-goal sizing approaches. The paper argues that the Skip-on-Fail schedule, which would otherwise degrade sample quality, is exactly what the two goal-selection mechanisms are designed to compensate for: PGDS keeps the curriculum near the policy's ability frontier, and CHER turns unsuccessful nominal-only transitions into conservative learning signals instead of discarding them. The reported outcome is a single policy that, at deployment, deterministically proposes device parameters for unseen target specifications and meets them at high success rates across a 45-corner evaluation grid.

Load-bearing premise

The approach assumes that a nominal-corner pass/fail gate plus a single per-specification worst-case value across the other corners carries enough information about PVT robustness to train a reliable policy.

Editorial extensions

If this is right

  • With the same number of environment steps, PPAAS meets target specifications in a larger fraction of evaluation episodes than the tested baselines on the TSA, CMA, and LDO benchmarks.
  • Because full-corner simulations are skipped whenever the nominal corner fails, the total number of simulator calls drops, with the reported ~4.1x simulation-efficiency gain averaged across benchmarks.
  • The trained goal-conditioned policy can be deployed on unseen target goals with deterministic mean actions, so new specifications need only tens of simulations rather than retraining.
  • The ablation results indicate that removing either PGDS or CHER lowers success rate by more than 8% on TSA, and removing the conservative reward computation widens seed-to-seed variance.
  • On the comparator benchmark, PGDS provides no success-rate gain over the baseline, which the paper attributes to the very low goal dimensionality.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If PGDS transfers, then any dense-reward goal-conditioned problem with a Pareto-ordered goal space could benefit from the same automatic curriculum, not just analog sizing.
  • The nominal gate in SoF is only as good as the assumption that a nominal failure implies nothing useful is learned about corner robustness; goals where nominal is the bottleneck would receive systematically weaker supervision.
  • The per-column worst-case aggregation suggests a testable extension: keeping per-corner vectors or explicit failure masks could handle corner interactions that a column-wise worst case misses.
  • The reported gains depend on comparisons against methods that run full-corner simulations or prune corners with clustering; under a fully parallel simulator the runtime advantage may shrink, though the sample-efficiency advantage should remain.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. This paper proposes PPAAS, a goal-conditioned Soft Actor-Critic framework for PVT-aware analog device sizing. The method has three main components: Skip-on-Fail (SoF), which runs full-corner simulations only after nominal-corner success; Pareto-goal sampling, which draws training goals from the non-dominated frontier of previously achieved goals using Q-value-based softmax weighting; and Conservative Hindsight Experience Replay (CHER), which relabels achieved goals with a conservative reward. The paper evaluates the method on four open-PDK benchmarks (TSA, CMA, LDO, COMP) and reports success rates up to 92.6%, with claimed average improvements of ~1.6x in sample efficiency and ~4.1x in simulation efficiency over prior methods. An ablation study on TSA supports the individual contributions of PGDS and CHER.

Significance. If the reported results survive scrutiny, PPAAS would be a practically useful step toward reducing the compute burden of PVT-robust analog sizing while preserving multi-goal generalization. The paper has clear strengths: public code, open-source PDKs and Ngspice, ten-seed averaging, and ablations that show each proposed component contributes to the final success rate. At the same time, the headline efficiency claims rest on a worst-case aggregation assumption in the reward and on ad hoc baseline adjustments, and the main comparison table does not report variance. The significance of the contribution is therefore real but conditional on addressing those load-bearing issues.

major comments (4)
  1. [Section III-B and III-D, Eq. (3)-(4)] The column-wise worst-case vector Worst(Zt) in Eq. (3) is not a sufficient statistic for multi-corner specification satisfaction. When different corners fail on different metrics, the worst-case vector can satisfy all targets even though no individual corner does; Eq. (4) then assigns a near-maximum dense reward in stage 2 despite D_t=0. Because the same collapsed achieved goal gt is also used in CHER relabeling (Algorithm 2, line 16), the optimistic signal propagates to relabeled transitions. The paper should either justify the sufficient-statistic assumption, replace z_t in ψ(·,·) with a per-corner minimum (or another aggregation that is conservative at the corner level), or provide an experiment measuring how often high-reward training transitions have D_t=0. The paper should also state explicitly that D_t is computed from the full matrix Zt and not from Worst(Zt).
  2. [Section IV-C2, Table III] The headline improvement ratios are computed against baselines with ad hoc adjustments: non-PVT methods are trained with a 20% expanded goal range and deployed on 20% harder goals, and RobustAnalog's COMP result uses a manually selected 8-corner subset rather than the full 16 corners, with footnote a indicating that the success rate is 0 otherwise. No sensitivity analysis is given for the 20% margin, and the COMP entry shows that the baseline comparison collapses without the manual corner selection. These choices are load-bearing for the claimed ~1.6x and ~4.1x improvements; please report unadjusted baseline results and vary the 20% margin to show the robustness of the improvement ratios.
  3. [Table III and Section IV-B] The text states that all metrics are means over ten random seeds, but Table III reports only point estimates, with no standard deviations or confidence intervals for SR, S_sim, or S_dev. Given the visible variance in the ablation curves (Fig. 4a) and the small sample-efficiency margins for some benchmarks, the current presentation does not establish whether the differences against prior methods are statistically meaningful. Please provide per-method mean ± std, or confidence intervals, for all entries in Table III.
  4. [Section IV-B, S_sim definition] The simulation-efficiency score S_sim = SR / #sim × 10^6 is partly mechanical: Skip-on-Fail reduces #sim by construction, so much of the reported ~4.1x simulation-efficiency gain is a direct consequence of the simulation-skipping strategy rather than of the goal-sampling or replay contributions. The paper should decompose the improvement into (a) the #sim reduction attributable to SoF and (b) the SR improvement attributable to PGDS/CHER, and state clearly which component drives the headline ratio.
minor comments (5)
  1. [Abstract and Section III-C] The acronym used in Algorithm 1 is 'Pareto Dominance Goal Sampling' (PDGS), but the abstract and elsewhere refer to 'Pareto-front Dominance Goal Sampling'; please unify the terminology.
  2. [Section IV-C2] The text says 'highest success rate and simulation efficiency on the TSO, CMA, and LDO benchmarks'; 'TSO' should be 'TSA'.
  3. [Abstract] The abstract says code is available 'HERE' rather than giving the actual URL; please replace the placeholder with the repository link.
  4. [Eq. (5)] The sensitivity of the reward normalizer η is not discussed; a sentence on how η=0.1 was chosen and whether results are stable to it would help reproducibility.
  5. [Section IV-A] The sentence describing the LDO initial state says s0 is initialized to the optimized values from [29], but it is not stated how those values are obtained or whether they are included in the public benchmark; please clarify.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity; results are empirical against external baselines and the self-citations are not load-bearing.

full rationale

PPAAS makes no claim of deriving a prediction from first principles; its central claims are measured success rates and simulation counts on four benchmark circuits implemented in open-source PDKs (GF180MCU/SKY130) with the open-source Ngspice simulator. The objective (Eq. 4), the achieved-goal construction (Eq. 3), PGDS, and CHER define the training mechanism, but the reported success rate is evaluated on uniformly pre-sampled held-out goals (Section IV-B), so the result does not reduce to the training signal by construction. The Worst(Zt) column-wise collapse in Eq. (3) is a modeling assumption that could be optimistic when different corners fail on different metrics, but this is a correctness risk, not a circular step. The only self-citations (RobustAnalog [9], DNN-Opt [12], CRONuS [6]) are baselines or comparisons used to position the method; they do not supply a load-bearing premise or a uniqueness theorem, and no result is imported from same-author prior work as an external proof. Skip-on-Fail directly reduces #sim and thereby raises Ssim by construction, but Ssim is explicitly defined as SR/#sim and the paper reports it as the intended mechanism rather than as an independent discovery, so this is honest reporting, not a renamed prediction. Accordingly, the circularity score is 1, reflecting only minor non-load-bearing self-citations.

Assumptions & free parameters 7 free parameters · 5 assumptions · 0 invented entities

The core empirical claim depends on simulator fidelity, corner set selection, worst-case goal aggregation, reward hyperparameters, and the distribution shift between PGDS training goals and uniform evaluation goals. No new physical entities are introduced.

free parameters (7)
  • Reward anchors R, Rmin, Rprime, Rmax = R=-1, Rmin=-6, Rprime=-3, Rmax=30
    These constants define the three-stage reward and the conservative relabeling; not derived from data.
  • PVT consistency weight alpha = 0 or 10
    Chosen to balance success rate versus Sdev; alpha=0 is default, alpha=10 is the robust variant.
  • Reward normalizer eta = 0.1
    Controls tanh sensitivity in Eq. 5; fixed by hand and not swept.
  • Goal sampling temperature T = 5.0
    Softmax temperature in PGDS; ablation in Fig. 4b shows performance depends on T.
  • Goal sampling candidates Ng and uniform threshold Nuniform = Ng=16, Nuniform=4
    Determines when and how many candidate goals PGDS considers.
  • SAC hyperparameter block = lr=0.003, batch=256, gamma=0.8, H=30
    Fixed training hyperparameters; no sweep reported for these.
  • Initial state s0 = best of 50 random candidates; LDO from [29]
    The starting design affects curriculum and final success; chosen without a principled rule.
assumptions (5)
  • domain assumption Ngspice and the open PDKs (GF180MCU, SKY130) produce simulator outputs that are faithful to real circuit behavior.
    All success rates and reward signals are computed from these simulations; no silicon measurements are reported.
  • domain assumption Training on 17 PVT corners transfers to deployment across 45 PVT corners.
    The policy is trained on a subset of corners and evaluated on a superset; transfer is assumed, not proven by the RL theory.
  • ad hoc to paper Worst-case per-metric aggregation across non-nominal corners is a sufficient statistic for multi-corner specification satisfaction.
    Eq. 3 defines gt with Worst(Zt) and the binary [Dnom,D], so corner-specific failure patterns are discarded.
  • ad hoc to paper The Pareto-front goal curriculum transfers to the uniformly sampled evaluation goal distribution.
    Training uses PGDS while evaluation samples targets uniformly; the distribution shift is expected to generalize but not guaranteed.
  • ad hoc to paper Dense rewards with the chosen tanh normalizer eta=0.1 guide optimization without reward hacking.
    The reward shape is an assumption; no robustness analysis is given for eta.

how reviews work

0 comments
Cite this review

Pith. "Pith review of PPAAS: PVT and Pareto Aware Analog Sizing via Goal-conditioned Reinforcement Learning." pith.science (2026). https://pith.science/paper/SYZKRWF6

@misc{pith2026250717003,
  author       = {Pith},
  title        = {Pith review of: PPAAS: PVT and Pareto Aware Analog Sizing via Goal-conditioned Reinforcement Learning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/SYZKRWF6}},
  note         = {Machine review of arXiv:2507.17003}
}
abstract

Device sizing is a critical yet challenging step in analog and mixed-signal circuit design, requiring careful optimization to meet diverse performance specifications. This challenge is further amplified under process, voltage, and temperature (PVT) variations, which cause circuit behavior to shift across different corners. While reinforcement learning (RL) has shown promise in automating sizing for fixed targets, training a generalized policy that can adapt to a wide range of design specifications under PVT variations requires much more training samples and resources. To address these challenges, we propose a \textbf{Goal-conditioned RL framework} that enables efficient policy training for analog device sizing across PVT corners, with strong generalization capability. To improve sample efficiency, we introduce Pareto-front Dominance Goal Sampling, which constructs an automatic curriculum by sampling goals from the Pareto frontier of previously achieved goals. This strategy is further enhanced by integrating Conservative Hindsight Experience Replay to stabilize training and accelerate convergence. To reduce simulation overhead, our framework incorporates a Skip-on-Fail simulation strategy. Experiments on benchmark circuits demonstrate $\sim$1.6$\times$ improvement in sample efficiency and $\sim$4.1$\times$ improvement in simulation efficiency compared to existing sizing methods. Code and benchmarks are publicly available at https://github.com/SeunggeunKimkr/PPAAS

Figures

Figures reproduced from arXiv: 2507.17003 by the authors.

Figure 1
Figure 1. Key differences between Single-goal RL and GCRL. GCRL is [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Overview of the proposed training process: It iteratively samples a goal using Pareto-Dominant Goal Sampling, rolls out an episode [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. Topologies of the four benchmark circuits. [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Ablation study on the TSA benchmark. The shaded region [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: Training runtime for each benchmark using 1, 2, 4, 8, and 16 [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. ORACLE: A Multi-Objective Reinforcement Learning-Based Analog Circuit Design Optimizer with Large Language Models-Guided Exploration

    eess.SY 2026-08 reject novelty 5.0 of 10

    ORACLE combines preference-conditioned multi-objective DDQN with LLM-guided action masking to size analog circuits, reporting high pass rates and large runtime cuts, but the evaluation compares 10 solutions per target...

Reference graph

Works this paper leans on

30 extracted references · 25 canonical work pages · cited by 1 Pith paper

  1. [1]

    Extraction and Use of Neural Network Models in Automated Synthesis of Operational Amplifiers,

    G. Wolfe and R. Vemuri, “Extraction and Use of Neural Network Models in Automated Synthesis of Operational Amplifiers,” IEEE TCAS I, 2003

    work page 2003

  2. [2]

    Batch Bayesian Optimization via Multi-objective Acquisition Ensemble for Automated Analog Circuit Design,

    W. Lyu, F. Yang, C. Yan, D. Zhou, and X. Zeng, “Batch Bayesian Optimization via Multi-objective Acquisition Ensemble for Automated Analog Circuit Design,” in Proc. ICML. PMLR, 2018, pp. 3306–3314

    work page 2018

  3. [3]

    Automated Design of Complex Analog Circuits with Multiagent Based Reinforcement Learning,

    J. Zhang, J. Bao, Z. Huang, X. Zeng, and Y . Lu, “Automated Design of Complex Analog Circuits with Multiagent Based Reinforcement Learning,” in Proc. DAC, 2023

    work page 2023

  4. [4]

    EVDMARL: Efficient Value Decomposition-Based Multi-Agent Reinforcement Learning with Domain-Randomization for Complex Analog Circuit Design Migration,

    H. Sun, Z. Bi, W. Jiang, Y . Lu, C. Yan, F. Yang, W. Hu, S.-G. Wang, D. Zhou, and X. Zeng, “EVDMARL: Efficient Value Decomposition-Based Multi-Agent Reinforcement Learning with Domain-Randomization for Complex Analog Circuit Design Migration,” in Proc. DAC, 2024

    work page 2024

  5. [5]

    Parasitic-Aware Analog Circuit Sizing with Graph Neural Networks and Bayesian Optimization,

    M. Liu, W. J. Turner, G. F. Kokai, B. Khailany, D. Z. Pan, and H. Ren, “Parasitic-Aware Analog Circuit Sizing with Graph Neural Networks and Bayesian Optimization,” in Proc. DATE, 2021

    work page 2021

  6. [6]

    CRONuS: Circuit Rapid Optimization with Neural Simulator,

    Y . Oh, D. Kim, Y . Lee, and B. Hwang, “CRONuS: Circuit Rapid Optimization with Neural Simulator,” in Proc. DATE, 2024

    work page 2024

  7. [7]

    PVTSizing: A TuRBO-RL-Based Batch-Sampling Optimization Framework for PVT- Robust Analog Circuit Synthesis,

    Z. Kong, X. Tang, W. Shi, Y . Du, Y . Lin, and Y . Wang, “PVTSizing: A TuRBO-RL-Based Batch-Sampling Optimization Framework for PVT- Robust Analog Circuit Synthesis,” in Proc. DAC, 2024

    work page 2024

  8. [8]

    Trust-Region Method with Deep Reinforcement Learning in Analog Design Space Exploration,

    K.-E. Yang, C.-Y . Tsai, H.-H. Shen, C.-F. Chiang, F.-M. Tsai, C.-A. Wang, Y . Ting, C.-S. Yeh, and C.-T. Lai, “Trust-Region Method with Deep Reinforcement Learning in Analog Design Space Exploration,” in Proc. DAC, 2021

    work page 2021

Show all 30 references
  1. [9]

    RobustAnalog: Fast Variation-Aware Analog Circuit Design Via Multi- task RL,

    W. Shi, H. Wang, J. Gu, M. Liu, D. Pan, S. Han, and N. Sun, “RobustAnalog: Fast Variation-Aware Analog Circuit Design Via Multi- task RL,” arXiv preprint arXiv:2207.06412 , 2022

    2022 arXiv

  2. [10]

    RoSE-Opt: Robust and Efficient Analog Circuit Parameter Optimiza- tion with Knowledge-infused Reinforcement Learning,

    W. Cao, J. Gao, T. Ma, R. Ma, M. Benosman, and X. Zhang, “RoSE-Opt: Robust and Efficient Analog Circuit Parameter Optimiza- tion with Knowledge-infused Reinforcement Learning,” arXiv preprint arXiv:2407.19150, 2024

    2024 arXiv

  3. [11]

    GCN-RL Circuit Designer: Transferable Transistor Sizing with Graph Neural Networks and Reinforcement Learning,

    H. Wang, K. Wang, J. Yang, L. Shen, N. Sun, H.-S. Lee, and S. Han, “GCN-RL Circuit Designer: Transferable Transistor Sizing with Graph Neural Networks and Reinforcement Learning,” in Proc. DAC, 2020

    2020

  4. [12]

    DNN-Opt: An RL Inspired Optimization for Analog Circuit Sizing Using Deep Neural Networks,

    A. F. Budak, P. Bhansali, B. Liu, N. Sun, D. Z. Pan, and C. V . Kashyap, “DNN-Opt: An RL Inspired Optimization for Analog Circuit Sizing Using Deep Neural Networks,” in Proc. DAC, 2021, pp. 1219–1224

    2021

  5. [13]

    Reinforcement Learning-based Analog Circuit Optimizer Using gm/ID for Sizing,

    M. Choi, Y . Choi, K. Lee, and S. Kang, “Reinforcement Learning-based Analog Circuit Optimizer Using gm/ID for Sizing,” in Proc. DAC, 2023

    2023

  6. [14]

    Automated Design of Analog Circuits Using Reinforcement Learning,

    K. Settaluri, Z. Liu, R. Khurana, A. Mirhaj, R. Jain, and B. Nikolic, “Automated Design of Analog Circuits Using Reinforcement Learning,” IEEE TCAD, 2022

    2022

  7. [15]

    Proxi- mal Policy Optimization Algorithms,

    J. Schulman, F. Wolski, P. Dhariwal, A. Radford, and O. Klimov, “Proxi- mal Policy Optimization Algorithms,” arXiv preprint arXiv:1707.06347, 2017

    2017 arXiv

  8. [16]

    Process Variations and Their Impact on Circuit Operation,

    S. Natarajan, M. Breuer, and S. Gupta, “Process Variations and Their Impact on Circuit Operation,” in Proceedings 1998 IEEE International Symposium on Defect and Fault Tolerance in VLSI Systems , 1998

    1998

  9. [17]

    A Sub- threshold Low-Power CMOS LC-VCO with High Immunity to PVT Variations,

    I. Ghorbel, F. Haddad, W. Rahajandraibe, and M. Loulou, “A Sub- threshold Low-Power CMOS LC-VCO with High Immunity to PVT Variations,” Analog Integr. Circuits Signal Process. , 2017

    2017

  10. [18]

    Universal Value Function Approximators,

    T. Schaul, D. Horgan, K. Gregor, and D. Silver, “Universal Value Function Approximators,” in Proc. ICML, 2015

    2015

  11. [19]

    Hindsight Experience Replay,

    M. Andrychowicz, F. Wolski, A. Ray, J. Schneider, R. Fong, P. Welinder, B. McGrew, J. Tobin, P. Abbeel, and W. Zaremba, “Hindsight Experience Replay,” in Proc. NIPS, 2018

    2018

  12. [20]

    AutoCkt: Deep Reinforcement Learning of Analog Circuit Designs,

    K. Settaluri, A. Haj-Ali, Q. Huang, K. Hakhamaneshi, and B. Nikolic, “AutoCkt: Deep Reinforcement Learning of Analog Circuit Designs,” in Proc. DATE, 2020

    2020

  13. [21]

    Curriculum learning,

    Y . Bengio, J. Louradour, R. Collobert, and J. Weston, “Curriculum learning,” in Proc. ICML, 2009

    2009

  14. [22]

    Automatic curriculum learning through value disagreement,

    Y . Zhang, P. Abbeel, and L. Pinto, “Automatic curriculum learning through value disagreement,” in Proc. NIPS, 2020

    2020

  15. [23]

    Quasimetric value functions with dense rewards,

    K. Valieva and B. Banerjee, “Quasimetric value functions with dense rewards,” arXiv preprint arXiv:2409.08724 , 2024

    2024 arXiv

  16. [24]

    Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor,

    T. Haarnoja, A. Zhou, P. Abbeel, and S. Levine, “Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor,” in Proc. ICML. PMLR, 2018, pp. 1861–1870

    2018

  17. [25]

    Continuous control with deep reinforcement learning,

    T. P. Lillicrap, J. J. Hunt, A. Pritzel, N. Heess, T. Erez, Y . Tassa, D. Silver, and D. Wierstra, “Continuous control with deep reinforcement learning,” arXiv preprint:1509.02971, 2019

    2019 arXiv

  18. [26]

    CROP: Conservative Reward for Model-based Offline Policy Optimization,

    H. Li, X.-H. Zhou, X.-L. Xie, S.-Q. Liu, Z.-Q. Feng, X.-Y . Liu, M.-J. Gui, T.-Y . Xiang, D.-X. Huang, B.-X. Yao, and Z.-G. Hou, “CROP: Conservative Reward for Model-based Offline Policy Optimization,” arXiv preprint arXiv:2310.17245 , 2023

    arXiv

  19. [27]

    AnalogGym: An Open and Practical Testing Suite for Analog Circuit Synthesis,

    J. Li, H. Zhi, R. Lyu, W. Li, Z. Bi, K. Zhu, Y . Zeng, W. Shan, C. Yan, F. Yang, Y . Li, and X. Zeng, “AnalogGym: An Open and Practical Testing Suite for Analog Circuit Synthesis,” arXiv preprint arXiv:2409.08534, 2024

    2024 arXiv

  20. [28]

    A Cascode Miller-Compensated Three-Stage Amplifier With Local Impedance Attenuation for Optimized Complex- Pole Control,

    M. Tan and W.-H. Ki, “A Cascode Miller-Compensated Three-Stage Amplifier With Local Impedance Attenuation for Optimized Complex- Pole Control,” IEEE Journal Solid-State Circuits , vol. 50, no. 2, pp. 440–449, 2015

    2015

  21. [29]

    Design and Optimization of Low-Dropout V oltage Regulator Using Relational Graph Neural Network and Rein- forcement Learning in Open-Source SKY130 Process,

    Z. Li and A. C. Carusone, “Design and Optimization of Low-Dropout V oltage Regulator Using Relational Graph Neural Network and Rein- forcement Learning in Open-Source SKY130 Process,” in Proc. ICCAD, 2023

    2023

  22. [30]

    [Online]

    GF180MCU PDK , GlobalFoundry. [Online]. Available: https://github. com/google/gf180mcu-pdk

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.