SPADE Zero Shot - EvoBench Overall TPR at 5 Percent FPR: leaderboard
The SPADE paper compares six zero-shot AI-text detectors using three released language-model backbones. AUROC and true-positive rate at 5% false positives remain separate. Paired base/instruct checkpoints are retained for Binoculars and Fast-DetectGPT. Learned soft prompts and supervised detector weights are distinct artifacts. This board reports tPR at 5 Percent FPR (%). Higher is better.
Metric: TPR at 5 Percent FPR (%). Source: arxiv.org. Saturation forecast: Rough model projection: around 2026. 18 models tracked. Updated 14 Nov 2026.
Top models
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | SPADE Binoculars: Llama 3.1 8B + Llama 3.1 8B Instruct | Evaluation run: frozen two-checkpoint detector | 56.9 | |
| 2 | SPADE Fast-DetectGPT: Llama 3.1 8B + Llama 3.1 8B Instruct | Evaluation run: frozen two-checkpoint detector | 55.6 | |
| 3 | SPADE Binoculars: Falcon-7B + Falcon-7B-Instruct | Evaluation run: frozen two-checkpoint detector | 52.3 | |
| 4 | SPADE Fast-DetectGPT: Falcon-7B + Falcon-7B-Instruct | Evaluation run: frozen two-checkpoint detector | 51.3 | |
| 5 | Qwen3-8B-Base (Harness: SPADE zero-shot Likelihood) | 49.3 | ||
| 6 | Qwen3-8B-Base (Harness: SPADE zero-shot Entropy) | 42.8 | ||
| 7 | Qwen3-8B-Base (Harness: SPADE zero-shot DNA-GPT) | 41.4 | ||
| 8 | Llama 3.1 8B (Harness: SPADE Zero Shot Likelihood) | 36.9 | ||
| 9 | falcon-7B (Harness: SPADE zero-shot DNA-GPT) | 35.1 | ||
| 10 | SPADE Binoculars: Qwen3-8B-Base + Qwen3-8B | Evaluation run: frozen two-checkpoint detector | 33.4 | |
| 11 | Llama 3.1 8B (Harness: SPADE Zero Shot DNA GPT) | 32.2 | ||
| 12 | SPADE Fast-DetectGPT: Qwen3-8B-Base + Qwen3-8B | Evaluation run: frozen two-checkpoint detector | 30.1 | |
| 13 | falcon-7B (Harness: SPADE zero-shot NTS) | 29.3 | ||
| 14 | falcon-7B (Harness: SPADE zero-shot Likelihood) | 29.1 | ||
| 15 | Llama 3.1 8B (Harness: SPADE Zero Shot Entropy) | 26.7 |
Interactive version: theaggregate.ai/benchmark?slug=spade-zero-shot-evobench-overall-tpr-at-5-percent-fpr · How It Works · Data refreshed daily, snapshot 2026-11-16.