ritwika96/qwen3.8-27b-arc1120-base-traces-20261025
Qwen base evaluation: 1,120 public ARC tasks Model: Qwen/Qwen3.8-27B, revision 1d4bf0f2ff6012fd82039f2fa52739d0dd7c60c0. No adapters. Thinking enabled. Four BF16 tensor-parallel replicas (two H200 GPUs each), with up to 32 concurrent solves per replica. Inventory: 1,000 public training tasks and 120 public evaluation tasks, 1,243 query grids. These are public-task diagnostic scores, not a private benchmark submission. Source hashes are in inventory.json. Each solve has 32,768… See the full description on the dataset page: https://huggingface.co/datasets/ritwika96/qwen3.8-27b-arc1120-base-trace
| Author / organization | ritwika96 |
|---|---|
| Downloads | 0 (as reported by Hugging Face) |
| Likes | 0 |
| Created | 25 October 2026 |
| Last updated at source | 25 October 2026 |
Recent changes
- 25 Oct 2026 Topics added arc-agi, reasoning, code-generation, evaluation
- 25 Oct 2026 Description: none → Qwen base evaluation: 1,120 public ARC tasks Model: Qwen/Qwen3.8-27B, revision 1d4bf0f2ff6012fd82039f2fa52739d0dd7c60c0. No adapters. Thinking enabled. Four BF16 tensor-parallel replicas (two H200 GPUs each), with up to 32 concurrent solves
Changes detected by Silytics between refreshes of the source.
More from ritwika96
Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 25 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.