A few days ago, we asked ChatGPT to write a composition for the college entrance examination .
Some people say that the writing is well-structured and well-organized, and they feel lucky that they graduated early, otherwise they would be surpassed by ChatGPT.
Some people think that the composition written by ChatGPT is quite satisfactory, bland, very correct but meaningless.
Today, we are going to let ChatGPT challenge the math questions of the college entrance examination again. The test questions are selected from the 2023 National New College Entrance Examination Volume I, and the scope of application includes candidates from Hubei, Shandong, Guangdong, Jiangsu, Hebei, Hunan, Fujian, Zhejiang and other places.
Since ChatGPT is temporarily unable to process the chart information , questions 10 and 18 were not given to ChatGPT to answer.
Now let's see how strong ChatGPT's ability to do math problems is.
1
multiple choice questions
There are 8 multiple-choice questions, each with 5 points, for a total of 40 points.
Among the four options given in each sub-item, only one item meets the requirements of the item.
Reference answer: C
Reference answer: A
Reference answer: D
Reference answer: D
Reference answer: A
Reference answer: B
Reference answer: C
Reference answer: B
In this part of the multiple choice questions, the 5th and 8th questions of ChatGPT were not solved, and all other questions were done correctly, with a total score of 30 points .
2
multiple choices
There are 4 multiple-choice questions, each with 5 points, for a total of 20 points. Among the options given in each sub-question, there are many that meet the requirements of the topic. 5 points for all correct answers, 2 points for partially correct answers, and 0 points for wrong answers.
Reference answer: BD
Reference answer: ABC
Reference answer: ABD
In the multiple-choice question section, ChatGPT did all the right things, except for the 10th question, which was not included, and the total score was 15 points .
3
fill in the blank
There are 4 small questions to fill in the blanks, each with 5 points, totaling 20 points.
Reference answer:
13 questions: 64
14 questions:
Question 15: 【2,3}
16 questions:
In the part of filling in the blanks, only the 14th question of ChatGPT was wrong, and the total score was 15 points .
4
answer question
There are 6 small questions to answer questions, with a total of 70 points. The answer should include the necessary text description, proof process or calculation steps.
Reference answer:
Reference answer:
Reference answer:
Reference answer:
Reference answer:
In this part of answering questions, ChatGPT only has 22 questions and sub-question 2 does not give the correct answer. If the score of this 1 small question is set to 6 points, then the total score is 52 points .
Knot
Now, let's summarize the overall performance of ChatGPT .
Excluding Question 10 (5 points) and Question 18 (12 points), the entire test paper is 133 points , and the ChatGPT score is 112 points .
Considering that ChatGPT is a large model based on English data , it is better at solving English problems.
Then when the whole set of test papers is 150 points, the points that ChatGPT can get can be added by 10 points.
Therefore, the total score of ChatGPT should be around 132 points .
According to previous college entrance examination results, students who can be admitted to 985 universities have an average score of around 130 in mathematics .
If ChatGPT is taking the college entrance examination in Guangdong , it can be admitted to Sun Yat-sen University according to its mathematics scores . If you are in Jiangsu, you can be admitted to Nanjing University ; if you are in Fujian, you can enter Xiamen University . If you are in Hunan, you can be admitted to Hunan University . If you are in Shandong, you can go to Shandong University ; if you are in Zhejiang, you can go to Zhejiang University .
Under the ChatGPT wave
5% of humans run ahead
95% of humans are still confused
but as long as you are willing to learn
ChatGPT is not that difficult
Today , Quantum School @ChatGPT
Write the first ChatGPT science book on carbon-silicon cooperation
" Silicon-Based Story: AI Big Bang "
Play the Rhapsody of Carbon-Silicon Collaboration
Use AI to change your life
From ChatGPT → AIGC → GPT-X →
AGI Evolution→ Magic Age →Human Future
From history to future, understand ChatGPT past and present
From theory to practice, use cases to analyze the specific operation of AIGC
From language to AI painting, practical exercises in midjourney
From technology to philosophy, exploring the possibility of AI life
In the new era of artificial intelligence
Hand in hand with AI
Become an " AIGC Transhuman Magician "
"Silicon-Based Story: AI Big Bang"
Winner of Wenjin Book Award Joint ChatGPT
Build a new era of carbon-silicon collaboration in the future
from 0 to 1
from 1 to 100
Read ChatGPT one day earlier
One day ahead of this era
Click on the applet below
▼ Now available for purchase ▼