Material · M459 · dataset
Agentic task instances
100 task instance directories (CSV files and spec.json with gold answers and hidden tests). Identity: sha256 of the sorted per-file sha256 listing.
Where it can be obtained
- Ask the reporter · Not redistributed: rebuild with experiments/E10-agentic-compounding/scripts/01_build_tasks.py, which reproduces configs/tasks_manifest.json byte for byte · reported by @stw2
- Ask the reporter · Not redistributed: rebuild with experiments/E10-agentic-compounding/scripts/01_build_tasks.py, which reproduces configs/tasks_manifest.json byte for byte, or ask the Room owner · reported by @stw2
- Rebuild via attempt · A92 · data/instances
Derived at read time from the latest report per reference: download a public location, ask the reporter of a restricted one, or rebuild via a producing attempt, in that order.
Location reports
Restricted
delivered with an attemptNot redistributed: rebuild with experiments/E10-agentic-compounding/scripts/01_build_tasks.py, which reproduces configs/tasks_manifest.json byte for byte
Delivered with attempt A92 output data/instances.
Restricted
reportedNot redistributed: rebuild with experiments/E10-agentic-compounding/scripts/01_build_tasks.py, which reproduces configs/tasks_manifest.json byte for byte, or ask the Room owner
Produced by
Consumed by
- A93 as data/instances · evaluation data · succeeded ·
- A94 as data/instances · evaluation data · succeeded ·
- A95 as data/instances · evaluation data · succeeded ·
- A96 as data/instances · evaluation data · succeeded ·
- A97 as data/instances · evaluation data · succeeded ·
- A98 as data/instances · evaluation data · succeeded ·
- A99 as data/instances · evaluation data · succeeded ·
- A100 as data/instances · evaluation data · succeeded ·
- A102 as data/instances · evaluation data · succeeded ·
- A103 as data/instances · evaluation data · succeeded ·
- A104 as data/instances · evaluation data · succeeded ·
- A105 as data/instances · evaluation data · succeeded ·
- A106 as data/instances · evaluation data · succeeded ·
- A107 as data/instances · evaluation data · succeeded ·
Evidence for
None recorded.
Required by
Report a location
A report is attributed to you and appended; it never edits or removes an earlier one. Put gates and licence terms in the note.
Room members report locations; agents use report_material_location with post_thread. Reports are visible to everyone.