Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2506.23719 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866913919369478144 |
|---|---|
| author | Egg, Alex Goyanes, Martin Iglesias Kingma, Friso Mora, Andreu von Werra, Leandro Wolf, Thomas |
| author_facet | Egg, Alex Goyanes, Martin Iglesias Kingma, Friso Mora, Andreu von Werra, Leandro Wolf, Thomas |
| contents | We introduce DABstep, a novel benchmark for evaluating AI agents on realistic multi-step data analysis tasks. DABstep comprises over 450 real-world challenges derived from a financial analytics platform, requiring models to combine code-based data processing with contextual reasoning over heterogeneous documentation. Each task demands an iterative, multi-step problem-solving approach, testing capabilities in data manipulation, cross-referencing multiple sources, and precise result reporting. The benchmark provides a factoid-style answer format with automatic correctness checks for objective scoring at scale. We evaluate leading LLM-based agents, revealing a substantial performance gap: even the best agent achieves only 14.55% accuracy on the hardest tasks. We detail our benchmark's design, dataset composition, task formulation, evaluation protocol, report baseline results and analyze failure modes. DABstep is released with a public leaderboard and toolkit to accelerate research in autonomous data analysis. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2506_23719 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | DABstep: Data Agent Benchmark for Multi-step Reasoning Egg, Alex Goyanes, Martin Iglesias Kingma, Friso Mora, Andreu von Werra, Leandro Wolf, Thomas Machine Learning Artificial Intelligence We introduce DABstep, a novel benchmark for evaluating AI agents on realistic multi-step data analysis tasks. DABstep comprises over 450 real-world challenges derived from a financial analytics platform, requiring models to combine code-based data processing with contextual reasoning over heterogeneous documentation. Each task demands an iterative, multi-step problem-solving approach, testing capabilities in data manipulation, cross-referencing multiple sources, and precise result reporting. The benchmark provides a factoid-style answer format with automatic correctness checks for objective scoring at scale. We evaluate leading LLM-based agents, revealing a substantial performance gap: even the best agent achieves only 14.55% accuracy on the hardest tasks. We detail our benchmark's design, dataset composition, task formulation, evaluation protocol, report baseline results and analyze failure modes. DABstep is released with a public leaderboard and toolkit to accelerate research in autonomous data analysis. |
| title | DABstep: Data Agent Benchmark for Multi-step Reasoning |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2506.23719 |