PARSE (Persian Zero-Shot) - Factoid Reasoning List: leaderboard

Metric: List match (0-1; answers containing a gold item over the larger of the gold and answer list sizes; 600 reasoning questions with two to five gold answers; Persian-language prompt, zero-shot, Together AI sampling). Source: arxiv.org. Saturation forecast: Estimated already saturated. 6 models tracked.

Top models

#ModelScore
1Gemma 2 27B (IT)0.28
2Llama 3.3 70B Instruct0.23
3Qwen 2.5 72B Instruct Turbo0.21
4Mistral Small 30.18

Interactive version: theaggregate.ai/benchmark?slug=parse-persian-zero-shot-factoid-reasoning-list · How It Works · Data refreshed daily, snapshot 2026-09-25.