GSM-Plus-BN - Problem Understanding: leaderboard
GSM-Plus-BN (2026) is a human-verified Bangla translation of the GSM-Plus perturbation benchmark for grade-school math. Accuracy in percent on the problem understanding variants (the problem reworded) with standard prompting; higher is better.
Metric: Accuracy (%; problem understanding perturbation, the problem reworded; Bangla grade-school math word problems translated from GSM-Plus and verified by human translators; standard prompting (answer directly); Groq API). Source: arxiv.org. Saturation forecast: Estimated already saturated. 6 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-OSS-120B | 89.82 | #700 |
| 2 | GPT-OSS-20B | 87.62 | #1147 |
| 3 | Llama 4 Scout Instruct | 87.43 | #1543 |
| 4 | Qwen 3 32B | 47.5 | #960 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=gsm-plus-bn-problem-understanding · How It Works · Data refreshed daily, snapshot 2026-11-13.