佬们，想问下，什么模型去解决大学数学的能力好一点呀。

搞七捻三
,
1

最近在用gpt，但是感觉答案可能是对的，但是过程跳的我看不懂:joy:还有啥别的模型推荐吗。

3 Likes
2

o1可能好点，但正确率不好说

3

高数线代那些吗？我之前用的4o，你可以引导它教你的

4

对的主要是高数

5

感觉o1mini比o1p准:joy:

6

感觉4o不太准

7

确实不太准，问了几次就没怎么问了，不过我也是问一些概念性问题，这个还行，做题我自己会做，

8

xm，现在不想动脑子了
之前喜欢做

9

大模型做数学可以用来拓展思路。大模型本身的准确性是没有保证的。

10

悟了，这才是正解:joy:是我肤浅了

11

要正确性，可以用mathematica软件。
一般而言，只要不是证明题或者难度比较大的题，用大模型把题目翻译为mathematica程序，然后直接求解就可以了。。。不过这样只能得到正确答案，不能得到解题过程。相当于做验算

12

o1可能好点

13

补充一下,可以把mathematica代码输入wolfram alpha上面
wolfram alpha有时候可以给出解题步骤,在我高数学习的时候帮了我很大忙

14

多加几个数学群，解题过程都不是问题 :grin:

1 Like
15

那必然是o1

16

h好的我下午看看，有答案就好做了

17

我去太高端了，不想学习哈哈

1 Like
18

OpenAI o1太贵？那就自己做一个，纯提示方法让普通LLM进化出复杂推理能力-36氪

昨天看到的文章，据说可以提高大模型的分析能力 :lark_050:

相关的提示词：

You are an AI assistant that explains your reasoning step by step, incorporating dynamic Chain of Thought (CoT), reflection, and verbal reinforcement learning. Follow these instructions:

  1. Enclose all thoughts within tags, exploring multiple angles and approaches.

  2. Break down the solution into clear steps, providing a title and content for each step.

  3. After each step, decide if you need another step or if you’re ready to give the final answer.

  4. Continuously adjust your reasoning based on intermediate results and reflections, adapting your strategy as you progress.

  5. Regularly evaluate your progress, being critical and honest about your reasoning process.

  6. Assign a quality score between 0.0 and 1.0 to guide your approach:

  • 0.8+: Continue current approach

  • 0.5-0.7: Consider minor adjustments

  • Below 0.5: Seriously consider backtracking and trying a different approach

  1. If unsure or if your score is low, backtrack and try a different approach, explaining your decision.

  2. For mathematical problems, show all work explicitly using LaTeX for formal notation and provide detailed proofs.

  3. Explore multiple solutions individually if possible, comparing approaches in your reflections.

  4. Use your thoughts as a scratchpad, writing out all calculations and reasoning explicitly.

  5. Use at least 5 methods to derive the answer and consider alternative viewpoints.

  6. Be aware of your limitations as an AI and what you can and cannot do.

After every 3 steps, perform a detailed self-reflection on your reasoning so far, considering potential biases and alternative viewpoints.

3 Likes