RTL-BenchLS - Repository-Issue Reasoning: leaderboard
Repository-issue reasoning on 108 real issue-fix pairs from open hardware repositories: given the repository context, the issue report and the buggy module, the model writes a fix that must match the developer patch, from RTL-BenchLS (2026), a benchmark of formally verified Verilog designs; the score is the functional pass rate in percent, where output must be formally equivalent to the golden design (Cadence LEC or SEC), and higher is better.
Metric: Functional pass rate (%) on repository-issue reasoning on 108 real issue-fix pairs: from the repository context, the issue report and the buggy design (the simplified setting with the bug location given) the model writes a fix that must be formally equivalent to the developer's patch; functional pass rate (%; output formally equivalent to the golden design, Cadence Conformal LEC for combinational and JasperGold SEC for sequential designs), temperature 0.2; higher is better. Source: arxiv.org. Saturation forecast: Around 2030. 8 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | DeepSeek V3.2 | 12 | #406 |
| 2 | Llama 3.3 70B Instruct | 8.3 | #1357 |
| 3 | GPT-4o | 6.3 | #741 |
| 4 | Claude Sonnet 4.5 | 5.6 | #250 |
| 5 | Qwen 3.5 397B A17B | 4.6 | #246 |
| 6 | GPT-4o Mini | 4 | #1623 |
| 7 | Claude Haiku 4.5 | 3.7 | #568 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=rtl-benchls-repository-issue-reasoning · How It Works · Data refreshed daily, snapshot 2026-11-13.