PARSE (Persian Zero-Shot) - Multiple-Choice Multihop Multi-Answer: leaderboard

Metric: Jaccard similarity of chosen and correct options (0-1; 600 four-option multi-hop questions with two to four correct options; Persian-language prompt, zero-shot, Together AI sampling). Source: arxiv.org. Saturation forecast: Estimated already saturated. 6 models tracked.

Top models

#ModelScore
1Llama 3.3 70B Instruct0.95
2Qwen 2.5 72B Instruct Turbo0.94
3Gemma 2 27B (IT)0.92
4Mistral Small 30.9

Interactive version: theaggregate.ai/benchmark?slug=parse-persian-zero-shot-multiple-choice-multihop-multi-answer · How It Works · Data refreshed daily, snapshot 2026-09-25.