00

ScaleRTL dataset

CodeArtifact ☆ WATCH ⎘ SHARE

First seen 9/3/2026

Last seen 9/3/2026

Evidence 4 chunks

01

NEIGHBORHOOD

6 nodes · 10 edges

graph · ScaleRTL dataset · depth=1

02

RELATIONSHIPS

5 connections

GRPO with State Mutations: Improving LLM-Based Hardware Test Plan Generation ← uses 100% 3e evidence

The paper uses the ScaleRTL dataset as the source of RTL codes for evaluation and training.

supervised fine-tuning ← uses 100% 1e evidence

SFT data is constructed from RTL designs drawn from the ScaleRTL dataset.

SFT training dataset ← derived from 100% 1e evidence

The SFT training dataset is constructed from RTL designs in the ScaleRTL dataset.

GRPO-SMu training dataset ← derived from 100% 1e evidence

The GRPO-SMu training dataset is constructed from golden RTL codes sourced from the ScaleRTL dataset.

ScaleRTL: scaling LLMs with reasoning data and test-time compute for accurate RTL code generation ← introduces 95% 1e evidence

The ScaleRTL paper introduces the ScaleRTL dataset used for RTL code generation.