3 days ago
-
Make formal-v1.sh parse under macOS's bash 3.2 and recognise a blocking web page dialog titled in English
-
Put the code comments and the audit variants' why.md in English (part 5)
-
Put the code comments and the audit variants' why.md in English (part 4)
-
Put the code comments and the audit variants' why.md in English (part 3)
-
Put the code comments and the audit variants' why.md in English (part 2)
-
Put the code comments and the audit variants' why.md in English
-
Put the judge audit notes in English
-
Add the tasks, judges, harness, app under test, docs and report; merge the repository README into the card (part 5)
-
Add the tasks, judges, harness, app under test, docs and report; merge the repository README into the card (part 4)
-
Add the tasks, judges, harness, app under test, docs and report; merge the repository README into the card (part 3)
-
Add the tasks, judges, harness, app under test, docs and report; merge the repository README into the card (part 2)
-
Add the tasks, judges, harness, app under test, docs and report; merge the repository README into the card
-
Cut the caption in the time-per-run figure to one line
-
Cut the text in the core-results figure to one line per place and centre the scoreboard digits
-
Lay out the failed-runs cell as a scoreboard: 4 and 8 side by side, names underneath
-
Label the failed-runs ratio "Mirasim" and "ChatGPT" so it fits on one line
-
Link Mirasim Computer Use's repository and npm package from the card
4 days ago
-
Add DevBrowse Bench v1 runs, tasks and dataset card
-
initial commit
