openai/simple-evals
File Size
The distribution of size of files (measured in lines of code).
File Size Overall
  • There are 18 files with 5,288 lines of code.
    • 1 very long files (2,535 lines of code)
    • 1 long files (528 lines of code)
    • 3 medium size files (990 lines of code)
    • 4 small files (590 lines of code)
    • 9 very small files (645 lines of code)
47% | 9% | 18% | 11% | 12%
Legend:
1001+
501-1000
201-500
101-200
1-100


explore: grouped by folders | grouped by size | sunburst | 3D view
File Size per Extension
1001+
501-1000
201-500
101-200
1-100
ipynb100% | 0% | 0% | 0% | 0%
py0% | 19% | 35% | 21% | 23%
File Size per Logical Decomposition
primary
1001+
501-1000
201-500
101-200
1-100
healthbench_scripts100% | 0% | 0% | 0% | 0%
ROOT0% | 21% | 40% | 24% | 13%
sampler0% | 0% | 0% | 0% | 100%
Longest Files (Top 18)
File# lines# units
healthbench_analysis.ipynb
in healthbench_scripts
 2535 -
528 16
496 1
272 8
222 20
common.py
in root
 194 12
146 6
134 1
116 2
91 5
89 3
88 5
86 5
80 5
65 5
63 2
49 2
types.py
in root
 34 2
Large Files That Change Often (Top 2)
To change a file, an AI coding agent (or a person) has to read it: these files are read the most for the changes in the past year (lines × changes).
File# lines≈ tokens
to read		# changes
(1y)		≈ lines edited
per change		lines read per
line changed
528 ~3k–10k 2 ~300 ~2
496 ~3k–7k 2 ~50 ~10

Tokens are lines × 7 to 14 (analysis.fileReadsForChanges). Lines edited per change are the file's lines added and deleted divided by its commits over the whole history. Agents read about 2,000 lines at a time: when a 17,000-line file was split into modules, the tokens an agent read for the same change fell by 83% (The Economic Benefit of Refactoring, 2026).

Files With Most Units (Top 17)
File# lines# units
222 20
528 16
common.py
in root
 194 12
272 8
146 6
86 5
91 5
88 5
80 5
65 5
89 3
49 2
63 2
types.py
in root
 34 2
116 2
496 1
134 1
Files With Long Lines (Top 5)

There are 5 files with lines longer than 120 characters. In total, there are 65 long lines.

File# lines# units# long lines
healthbench_analysis.ipynb
in healthbench_scripts
 2535 - 57
528 16 5
272 8 1
91 5 1
89 3 1
Correlations

File Size vs. Commits (all time): 18 points

healthbench_eval.py x: 4 commits (all time) y: 528 lines of code simple_evals.py x: 26 commits (all time) y: 496 lines of code healthbench_scripts/healthbench_analysis.ipynb x: 3 commits (all time) y: 2535 lines of code browsecomp_eval.py x: 8 commits (all time) y: 86 lines of code common.py x: 11 commits (all time) y: 194 lines of code drop_eval.py x: 5 commits (all time) y: 222 lines of code gpqa_eval.py x: 9 commits (all time) y: 63 lines of code healthbench_meta_eval.py x: 2 commits (all time) y: 272 lines of code math_eval.py x: 8 commits (all time) y: 49 lines of code mgsm_eval.py x: 8 commits (all time) y: 146 lines of code mmlu_eval.py x: 7 commits (all time) y: 116 lines of code sampler/chat_completion_sampler.py x: 3 commits (all time) y: 80 lines of code sampler/claude_sampler.py x: 3 commits (all time) y: 91 lines of code sampler/o_chat_completion_sampler.py x: 4 commits (all time) y: 65 lines of code sampler/responses_sampler.py x: 9 commits (all time) y: 88 lines of code simpleqa_eval.py x: 12 commits (all time) y: 89 lines of code types.py x: 3 commits (all time) y: 34 lines of code run_multilingual_mmlu.py x: 7 commits (all time) y: 134 lines of code
2535.0
lines of code
  min: 34.0
  average: 293.78
  25th percentile: 76.25
  median: 103.5
  75th percentile: 234.5
  max: 2535.0
0 26.0
commits (all time)
min: 2.0 | average: 7.33 | 25th percentile: 3.0 | median: 7.0 | 75th percentile: 9.0 | max: 26.0

File Size vs. Contributors (all time): 18 points

healthbench_eval.py x: 3 contributors (all time) y: 528 lines of code simple_evals.py x: 9 contributors (all time) y: 496 lines of code healthbench_scripts/healthbench_analysis.ipynb x: 2 contributors (all time) y: 2535 lines of code browsecomp_eval.py x: 4 contributors (all time) y: 86 lines of code common.py x: 7 contributors (all time) y: 194 lines of code drop_eval.py x: 5 contributors (all time) y: 222 lines of code gpqa_eval.py x: 8 contributors (all time) y: 63 lines of code healthbench_meta_eval.py x: 2 contributors (all time) y: 272 lines of code math_eval.py x: 7 contributors (all time) y: 49 lines of code mgsm_eval.py x: 7 contributors (all time) y: 146 lines of code mmlu_eval.py x: 6 contributors (all time) y: 116 lines of code sampler/chat_completion_sampler.py x: 3 contributors (all time) y: 80 lines of code sampler/claude_sampler.py x: 3 contributors (all time) y: 91 lines of code sampler/o_chat_completion_sampler.py x: 4 contributors (all time) y: 65 lines of code simpleqa_eval.py x: 7 contributors (all time) y: 89 lines of code types.py x: 3 contributors (all time) y: 34 lines of code run_multilingual_mmlu.py x: 3 contributors (all time) y: 134 lines of code
2535.0
lines of code
  min: 34.0
  average: 293.78
  25th percentile: 76.25
  median: 103.5
  75th percentile: 234.5
  max: 2535.0
0 9.0
contributors (all time)
min: 2.0 | average: 4.83 | 25th percentile: 3.0 | median: 4.0 | 75th percentile: 7.0 | max: 9.0

File Size vs. Commits (30 days): 0 points

No data for "commits (30d)" vs. "lines of code".

File Size vs. Contributors (30 days): 0 points

No data for "contributors (30d)" vs. "lines of code".

File Size vs. Commits (90 days): 0 points

No data for "commits (90d)" vs. "lines of code".

File Size vs. Contributors (90 days): 0 points

No data for "contributors (90d)" vs. "lines of code".

About This Analysis
  • File size measurements show the distribution of size of files.
  • Files are classified in four categories based on their size (lines of code): 1-100 (very small files), 101-200 (small files), 201-500 (medium size files), 501-1000 (long files), 1001+(very long files).
  • It is a good practice to keep files small. Long files may become "bloaters", code that have increased to such gargantuan proportions that they are hard to work with.
Learn more...