按年被提到
20242024 · 4 次2026
大模型与语言模型›大模型推理与思维链·arXiv:2405.00332 · cs.CL

A Careful Examination of Large Language Model Performance on Grade School Arithmetic

Hugh Zhang, Jeff Da, Dean Lee, Vaughn Robinson, Catherine Wu, Will Song, Tiffany Zhao, Pranav Raja, Charlotte Zhuang, Dylan Slack, Qin Lyu, Sean Hendryx 等 15 人 · 2024
开始阅读 ↗PDFalphaXiv 讨论
被提及 · 5 人首次 2024-05-02 · 最近 2025-12-02
5不同的人
2活跃月份
5讨论串
43被引(OpenAlex)
arXiv 摘要(候选层不写导读)

Large language models (LLMs) have achieved impressive success on many benchmarks for mathematical reasoning. However, there is growing concern that some of this performance actually reflects dataset contamination, where data closely resembling benchmark questions leaks into the training data, instead of true reasoning ability. To investigate this claim rigorously, we commission Grade School Math 1000 (GSM1k). GSM1k is designed to mirror the style and complexity of the established GSM8k benchmark, the gold standard for measuring elementary mathematical reasoning. We ensure that the two benchmarks are comparable across important metrics such as human solve rates, number of steps in solution, answer magnitude, and more. When evaluating leading open- and closed-source LLMs on GSM1k, we observe accuracy drops of up to 8%, with several families of models showing evidence of systematic overfitting across almost all model sizes. Further analysis suggests a positive relationship (Spearman's r^2 = 0.36) between a model's probability of generating an example from GSM8k and its performance gap between GSM8k and GSM1k, suggesting that some models may have partially memorized GSM8k. Nevertheless, many models, especially those on the frontier, show minimal signs of overfitting, and all models broadly demonstrate generalization to novel math problems guaranteed to not be in their training data.

候选层:至少 2 个不同的人、在 2 个不同月份提到过,还没到主书架的 3 人 3 个月。没有做立场标注,所以只写“被提及”。

HN 书评

推荐在前,同类按时间 · 每篇最多保留 20 条,统计用全部提及
评论kalkin · 2025-12

“Scale AI wrote a paper a year ago comparing various models performance on benchmarks to performance on similar but held-out questions. Generally the closed source models performed better, and Mistral came out looking pretty badly: https://arxiv.org/pdf/2405.00332”

Mistral 3 family of models released原评论 ↗
被提交为 story · 4 次