GSM-Plus-BN - Numerical Substitution: 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 numerical substitution variants (numbers replaced by other values of the same digit length) with standard prompting; higher is better.
Metric: Accuracy (%; numerical substitution perturbation, numbers replaced by other values of the same digit length; 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 | Llama 4 Scout Instruct | 85.23 | #1543 |
| 2 | GPT-OSS-120B | 84.03 | #700 |
| 3 | GPT-OSS-20B | 79.64 | #1147 |
| 4 | Qwen 3 32B | 42.91 | #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-numerical-substitution · How It Works · Data refreshed daily, snapshot 2026-11-13.