DPrivBench - Report Noisy Max (Laplace): leaderboard

Differential privacy verification for the report-noisy-max with Laplace noise on the mechanism-level instances of DPrivBench (2026), where the model decides whether a stated differential privacy guarantee holds for a textbook mechanism instantiated with a concrete query, no tools, five seeds. Accuracy in percent; higher is better.

Metric: Accuracy (0-1, times 100) on the report-noisy-max with Laplace noise instances of the DPrivBench Category 1 mechanism-level instances (a mechanism template instantiated with a concrete query function from a 49-function bank; positive instances calibrate the noise to the true sensitivity, paired negatives under-calibrate it); binary yes/no verdict on whether the stated differential privacy guarantee holds, five seeds, no tools; higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 10 models tracked. Updated 7 Oct 2026.

Top models

#ModelSettingScoreOverall rank
1GPT-5High99.8#185
2Gemini 3 Pro64.5#146
3DeepSeek V3.1Non-reasoning64.3#511
4Claude Opus 4.559#165
5Claude Sonnet 4.554.9#267
6Gemini 2.5 Flash51.4#466
7Qwen 3 30B A3B 2507 Instruct50.8#1106
8Goedel-Prover-V2-32B46.9
9Qwen 3 30B A3B 2507 (Thinking)41.6#769
10DeepSeek R135.9#484

No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=dprivbench-report-noisy-max-laplace · How It Works · Data refreshed daily, snapshot 2026-11-22.