Research map / Graph learning / Knowledge and temporal graphs
DynamicRTL: RTL Representation Learning for Dynamic Circuit Behavior
Ruiyang Ma, Yunhao Zhou, Yipeng Wang, Yi Liu, Zhengyuan Shi, Ziyang Zheng, Kexin Chen, Zhiqiang He, Lingwei Yan, Gang Chen, Qiang Xu, Guojie Luo
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add branch supervision to toggle-only pretraining (ablation) | DR-GNN · Assertion v1&v2==0 | accuracy (%): 82.85 -> 86.80 | Improved |
| add toggle supervision to branch-only pretraining (ablation) | DR-GNN · Assertion v1&v2==0 | accuracy (%): 66.80 -> 86.80 | Improved |
| add toggle supervision to branch-only pretraining (ablation) | DR-GNN · Assertion v1 < v2 | accuracy (%): 77.18 -> 88.72 | Improved |
| add branch supervision to toggle-only pretraining (ablation) | DR-GNN · Assertion v1 < v2 | accuracy (%): 81.67 -> 88.72 | Improved |
| use both supervisions instead of either alone (ablation) | DR-GNN · Assertion v != 4 | accuracy (%): 80.59 -> 81.36 | Improved |
| use both pretraining supervisions instead of one (ablation) | DR-GNN · Power estimation | accuracy | Improved |
| use binary input encoding instead of random or decimal encoding (ablation) | DR-GNN · Branch hit / toggle rate prediction | accuracy | Improved |
| use semi-decoupled aggregation instead of fully decoupled or shared (ablation) | DR-GNN · Branch hit / toggle rate prediction | accuracy | Improved |
| add operator type and operand position to attention messaging (ablation) | DR-GNN · Branch hit / toggle rate prediction | accuracy | Improved |
| use operand position embedding instead of Laplacian or distance position encodings (ablation) | DR-GNN · Branch hit / toggle rate prediction | accuracy | Improved |
| use RTL-level DR-GNN instead of netlist-level DeepSeq2 | power estimation · pcie_ctrl | error: 0.08 -> 0.15 | Hurt |
| use RTL-level DR-GNN instead of netlist-level DeepSeq2 | power estimation · dma_sched | error: 0.09 -> 0.14 | Hurt |
Closest papers by meaning
- Bridging Layout and RTL: Knowledge Distillation based Timing Prediction (ICML 2025 · similarity 0.85)
- Circuit Representation Learning with Masked Gate Modeling and Verilog-AIG Alignment (ICLR 2025 · similarity 0.84)
- DeepGate4: Efficient and Effective Representation Learning for Circuit Design at Scale (ICLR 2025 · similarity 0.84)
- Efficient Automated Circuit Discovery in Transformers using Contextual Decomposition (ICLR 2025 · similarity 0.84)
- CircuitFusion: Multimodal Circuit Representation Learning for Agile Chip Design (ICLR 2025 · similarity 0.83)
- Dynamic Spiking Graph Neural Networks (AAAI 2024 · similarity 0.83)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
