SearchAbout
#ai-researchStory

AgenticDataBench: A Comprehensive Benchmark for Data Agents

AgenticDataBench is a new benchmark for evaluating data agents across diverse domains with fine-grained task annotations and skill-based coverage metrics. It includes 344 tasks, 433 data science skills, 97 datasets, and 27.3GB of data spanning 15 domains. The benchmark enables both overall task success evaluation and fine-grained skill-level performance analysis.

#ai-research#benchmark#data-agents#data-science#llms
Hugging Face Daily Papers4 min read4d ago
AgenticDataBench: A Comprehensive Benchmark for Data Agents
Read original ↗Sign in to upvoteSign in to save
ShareXLinkedInHacker News

Related stories

MemSyco-Bench: Benchmarking Sycophancy in Agent MemoryHugging Face Daily Papers5d agoScarfBench: Benchmarking AI Agents for Enterprise Java Framework MigrationHugging Face Blog5d agoAgenticSTS: A Bounded-Memory Testbed for Long-Horizon LLM AgentsHugging Face Daily Papers4d agoAGVBench: A Reliability-Oriented Benchmark of Data Augmentation for Vein RecognitionHugging Face Daily Papers4d ago