Are you planning to release the answers to the remaining 400 questions?

[hyyp1]

I would like to use this dataset to test my model, but I found that 400 of the entries don't even have ground truths, which makes this part unusable for any evaluation.

[StephenQSstarThomas]

Thanks for your attention. Unfortunately, to keep our benchmark as a role of TRUE and RIGOROUS evaluation, we'll reserve the remaining answers of our questions. However, we're building an automatic pipeline of evaluation and it will be released in the next 2 weeks. I'll send the link and close this discussion at that time.

[joey9503]

怎么样啦？有网站可以测试剩下的400道题目吗？

[StephenQSstarThomas]

网站的测试接口会在近期发布（两周内），PHYBench团队大多是本科生（甚至是低年级本科生），最近期末季了请稍等

[GaoZhitao121]

网站的测试接口会在近期发布（两周内），PHYBench团队大多是本科生（甚至是低年级本科生），最近期末季了请稍等

请问有测试接口了吗，目前在网站里面有找到诶

[TIMMY-CHAN]

您好，请问剩下的400道题有测试接口了吗～

[StephenQSstarThomas]

We'll gonna release all questions and their evaluation in this month. Just wait for 2 weeks~

[SpiritsYouthHarmony]

We'll gonna release all questions and their evaluation in this month. Just wait for 2 weeks~

两个月过去了hhh，有消息了吗？👀

[StephenQSstarThomas]

Still finalizing 'cause we found some answer equivalance, with which we have to change the expression of the questions, to make the answers unique.