2D-FET-Bench V2
Harbor: pendingA benchmark of 128 layout tasks for evaluating language-model agents in generating geometrically valid FET designs on microscopy-derived 2D flake contours.
- Samples
- 128
- Released
- 2026
- Scores
- 6
CritPt
CritPt — AI evaluation benchmark listed by opencompass.
How to run
ltb · critpt
$ npm i -g lettokenburn-cli$ ltb bench pull critpt
External harness. pip install opencompass && opencompass --models <model_cfg> --datasets critpt_gen (OpenCompass dataset config dir 'CritPt') https://github.com/open-compass/opencompass
Runs happen on your machine with your own subscription; only the score summary is pushed. CLI docs
Scores
138 sourced entries| Harness | Source | Confidence | Link | |||
|---|---|---|---|---|---|---|
| GPT-5.6 Sol | 32.3 | N/A | Independent | low | N/A | epoch.ai |
| Claude Opus 5.5 | 31.7 | N/A | Independent | low | N/A | epoch.ai |
| GPT-6.1 Sol | 31.7 | N/A | Independent | low | N/A | epoch.ai |
| GPT-6 Astra | 31.7 | N/A | Independent | low | N/A | epoch.ai |
| Claude Sonnet 5.5 | 31.4 | N/A | Independent | low | N/A | epoch.ai |
| Claude Fable 5.1 | 31.1 | N/A | Independent | low | N/A | epoch.ai |
| GPT-6 Sol | 30.9 | N/A | Independent | low | N/A | epoch.ai |
| GPT-5.5 Pro | 30.6 | N/A | Independent | low | N/A | epoch.ai |
| GPT-5.4 Pro | 30.0 | N/A | Independent | low | N/A | epoch.ai |
| GPT-5.6-terra | 30.0 | N/A | Independent | low | N/A | epoch.ai |
| Claude Opus 5 | 29.1 | N/A | Independent | low | N/A | epoch.ai |
| Claude Fable 5 | 28.6 | N/A | Independent | low | N/A | epoch.ai |
| Gemini 4 Argon | 27.1 | N/A | Independent | low | N/A | epoch.ai |
| GPT-5.5 | 27.1 | N/A | Independent | low | N/A | epoch.ai |
| MiMo V2.6 Pro | 26.6 | N/A | Independent | low | N/A | epoch.ai |
| Gemini 3 Deep Think | 25.7 | N/A | Independent | low | N/A | epoch.ai |
| Muse Spark 1.3 | 24.9 | N/A | Independent | low | N/A | epoch.ai |
| GPT-5.4 | 23.4 | N/A | Independent | low | N/A | epoch.ai |
| Kimi K3 | 23.4 | N/A | Independent | low | N/A | epoch.ai |
| Claude Opus 4.8 | 20.9 | N/A | Independent | low | N/A | epoch.ai |
| GLM-5.2 | 20.9 | N/A | Independent | low | N/A | epoch.ai |
| Step 5 Preview | 20.9 | N/A | Independent | low | N/A | epoch.ai |
| GPT-5.6 Luna | 20.6 | N/A | Independent | low | N/A | epoch.ai |
| Qwen3.8 Max | 20.0 | N/A | Independent | low | N/A | epoch.ai |
| Grok-4.6 | 19.7 | N/A | Independent | low | N/A | epoch.ai |
| GPT-6 Luna | 19.4 | N/A | Independent | low | N/A | epoch.ai |
| GLM-5.3 | 19.1 | N/A | Independent | low | N/A | epoch.ai |
| Gemini 3.8 Flash | 18.3 | N/A | Independent | low | N/A | epoch.ai |
| DeepSeek V4 Pro 0813 | 18.0 | N/A | Independent | low | N/A | epoch.ai |
| Gemini 3.1 Pro | 17.7 | N/A | Independent | low | N/A | epoch.ai |
| Grok-4.7 | 17.7 | N/A | Independent | low | N/A | epoch.ai |
| Muse Spark 1.2 | 17.7 | N/A | Independent | low | N/A | epoch.ai |
| Qwen3.8 Max 0902 | 17.7 | N/A | Independent | low | N/A | epoch.ai |
| Claude Sonnet 5 | 16.9 | N/A | Independent | low | N/A | epoch.ai |
| DeepSeek V4 Flash 0731 | 16.6 | N/A | Independent | low | N/A | epoch.ai |
| GLM-5.3 Flash | 15.4 | N/A | Independent | low | N/A | epoch.ai |
| Grok 4.5 | 15.4 | N/A | Independent | low | N/A | epoch.ai |
| Muse Spark 1.1 | 15.1 | N/A | Independent | low | N/A | epoch.ai |
| DeepSeek V4.1 Flash | 14.3 | N/A | Independent | low | N/A | epoch.ai |
| Gemini 3.7 Flash | 14.3 | N/A | Independent | low | N/A | epoch.ai |
| Qwen3.7 Max | 13.4 | N/A | Independent | low | N/A | epoch.ai |
| Gemini-3.5-flash | 13.1 | N/A | Independent | low | N/A | epoch.ai |
| DeepSeek V4 Pro | 12.9 | N/A | Independent | low | N/A | epoch.ai |
| GPT-5 | 12.6 | N/A | Independent | low | N/A | epoch.ai |
| Claude Opus 4.7 | 12.0 | N/A | Independent | low | N/A | epoch.ai |
| MiMo V2.6 Flash | 12.0 | N/A | Independent | low | N/A | epoch.ai |
| Muse Spark | 11.3 | N/A | Independent | low | N/A | epoch.ai |
| Gemini 3.6 Flash | 10.6 | N/A | Independent | low | N/A | epoch.ai |
| GPT-5.4-mini | 10.0 | N/A | Independent | low | N/A | epoch.ai |
| Kimi K2.7 Code | 10.0 | N/A | Independent | low | N/A | epoch.ai |
Top 50 of 138 by score
Every number links to where it was published. Low-confidence rows are muted and need per-cell verification.
Reproduced by the community
All runsLoading community runs…
About
Tracked by Epoch AI's Benchmarking Hub (results collected from external sources). Random baseline 0; score ceiling 1 (0–1 scale).