What Limits Virtual Agent Application? OmniBench: A Scalable Multi-Dimensional Benchmark for Essential Virtual Agent Capabilities
This work addresses the problem of inadequate benchmarking for virtual agents, which is crucial for researchers and developers in AI, though it is incremental as it builds on existing MLLM-based agent frameworks.
The authors tackled the limitations of existing benchmarks for virtual agents by introducing OmniBench, a scalable, self-generating benchmark with 36k graph-structured tasks across 20 scenarios, achieving a 91% human acceptance rate, and OmniEval, a multidimensional evaluation framework that reveals performance across 10 capabilities for various models.
As multimodal large language models (MLLMs) advance, MLLM-based virtual agents have demonstrated remarkable performance. However, existing benchmarks face significant limitations, including uncontrollable task complexity, extensive manual annotation with limited scenarios, and a lack of multidimensional evaluation. In response to these challenges, we introduce OmniBench, a self-generating, cross-platform, graph-based benchmark with an automated pipeline for synthesizing tasks of controllable complexity through subtask composition. To evaluate the diverse capabilities of virtual agents on the graph, we further present OmniEval, a multidimensional evaluation framework that includes subtask-level evaluation, graph-based metrics, and comprehensive tests across 10 capabilities. Our synthesized dataset contains 36k graph-structured tasks across 20 scenarios, achieving a 91\% human acceptance rate. Training on our graph-structured data shows that it can more efficiently guide agents compared to manually annotated data. We conduct multidimensional evaluations for various open-source and closed-source models, revealing their performance across various capabilities and paving the way for future advancements. Our project is available at https://omni-bench.github.io/.