FrontierCO is a curated benchmark suite for evaluating machine learning-based solvers on large-scale and real-world combinatorial optimization problems. The benchmark spans 8 classical CO problems across 5 application domains, providing both training and evaluation data. It was created by CO-Bench and last updated on Hugging Face in January 2026.
Use Cases
- Benchmarking language model agents on algorithm search for combinatorial optimization problems.
- Training ML-based solvers on large-scale combinatorial optimization problems.
- Evaluating solver performance across 8 classical CO problems in 5 application domains.
- Comparing algorithm performance on real-world combinatorial optimization instances.
Strengths
- Covers 8 classical combinatorial optimization problems.
- Spans 5 distinct application domains.
- Provides both training and evaluation data.
Limitations
- Column-level documentation is absent; field semantics must be inferred after download.
- Row count is unknown, which may limit suitability assessment.
- Description metadata is limited; actual data quality requires manual inspection after download.
Provenance
- Source
- CO-Bench
- Collection Method
- Curated benchmark suite.
- Time Range
- null
- Freshness
- Last updated 2026-01-12 05:56:39; freshness should be verified.
- Geography
- null