Launched the first nationwide new question bank on 7.29.Practice toolSince then, I have been continuously improving the coverage of solution explanations with AI assistance.
We have already run a trial with the following models: Tongyi Qianwen 3-Plus (with RAG), GPT-5-Chat, and Claude-4-Sonnet, on a set of A-level questions. We also performed manual review, and found that the following A-level questions had issues:
MC2-0035, Analysis:https://exam.bh5hsu.com/question/MC2-0035Item C has clear legal provisions that support its correctness.
MC4-0066, Option D, has a limit of 1W EIRP in the LF band at 136 kHz. Stating that LF applications less than 15W are allowed may be controversial.
MC2-0167, Analysis:https://exam.bh5hsu.com/question/MC2-0167Item C has clear legal provisions that support its correctness.
MC1-0185, in the question stem, WARC-76 should be WARC-79.
MC1-0437, option D should be correct. Option A contains inaccuracies (it needs to specify that it is impossible to receive on a single frequency simultaneously).
MC1-1300, with a 50% duty cycle square wave signal, the effective value is not 1/2 of the maximum value. Please see the explanation below:https://exam.bh5hsu.com/question/MC1-1300
MC1-1365, according to GB/T 3805-2008, the DC voltage limit in a humid state is 35V, not 33V as specified in the options.
Furthermore, some previously known problematic questions from older question banks were not updated.
MC1-1008, Analysis:https://exam.bh5hsu.com/question/MC1-1008
MC1-1009, Analysis:https://exam.bh5hsu.com/question/MC1-1009
Are these questions actually wrong? What other mistakes have you found?
闽ICP备2021006864号-7
闽公网安备35020602002794号