AuroraSystem/ABench75
ABench75 A compact bilingual (Russian + English) benchmark for instruction-following and short-answer reasoning across five domains: math, logic, code, cybersecurity, general. Total tasks: 375 (75 per domain) Public split: 300 tasks (data/tasks_public.jsonl) Private holdout: 75 tasks (15 per domain, not published) — for honest tracking Languages: ~59% English / ~41% Russian Scoring: regex match on the final answer (re field, case-insensitive) Canary: ABENCH75_CANARY_8f3c1d9a.… See the full description on the dataset page: https://huggingface.co/datasets/AuroraSystem/ABench75.
| Author / organization | AuroraSystem |
|---|---|
| License (as declared) | cc-by-4.0 |
| Downloads | 0 (as reported by Hugging Face) |
| Likes | 0 |
| Created | 20 October 2026 |
| Last updated at source | 20 October 2026 |
Recent changes
- 20 Oct 2026 License: none → cc-by-4.0
- 20 Oct 2026 Description: none → ABench75 A compact bilingual (Russian + English) benchmark for instruction-following and short-answer reasoning across five domains: math, logic, code, cybersecurity, general. Total tasks: 375 (75 per domain) Public split: 300 tasks (data/t
- 20 Oct 2026 Topics added task_categories:question-answering, task_categories:text-generation, benchmark, evaluation, reasoning, cybersecurity, math, logic, code
Changes detected by Silytics between refreshes of the source.
Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 20 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.