human_eval_plus: by examples

Results Paper Code

Not solved by any model

There are 10 examples not solved by any model. Solving some of these can be a good signal that your model is indeed better than leading models if these are good problems.
130, 139, 140, 145, 28, 32, 4, 51, 83, 86

Problems solved by 1 model only

example_link	model	min_pass1_of_model

Suspect problems

These are 10 problems with the lowest correlation with the overall evaluation (i.e. better models tend to do worse on these. )

example_link	pass1_of_ex	tau
21	0.014	-0.134
154	0.125	-0.034
163	0.001	0.013
22	0.052	0.073
127	0.121	0.176
132	0.001	0.234
48	0.310	0.287
55	0.271	0.302
35	0.787	0.312
54	0.404	0.325

Histogram of accuracies

Histogram of problems by the accuracy on each problem.

Histogram of difficulties

Histogram of problems by the minimum win rate to solve each problem.