nano-gpt-data-curation
gpt-5.6-sol
nano-gpt-data-curation — graded by the task's verifier against a measured baseline.
- backend
- modal
- variant
- codex azure_ai/gpt-5.6-sol
- metric
- held-out multi-domain perplexity
- actions
- 218
- duration
- 33m
- recorded
- 2026-08-04
run iddclm-nano-curation__9upVhik
Score
reward
333.760
Summary
Completed all deliverables:
- selection.json: 11,309 unique valid IDs; verified to pack exactly 12M tokens.
- curate.py: reproducible contrastive ranking, filtering, deduplication, and domain balancing.
- claim.md: hypothesis, mechanism, falsification, and transfer.
Frozen dev evaluation:
- Curated perplexity: 326.790
- Random baseline test: 476.714
- Improvement: 31.4% lower perplexity.