Beyond Perfect APIs: A Comprehensive Evaluation of LLM Agents Under Real-World API Complexity
This addresses the gap in evaluating LLM agents for real-world API interactions, though it is incremental as it builds on prior work by adding realistic complexity factors.
The authors tackled the problem of evaluating LLM agents' function-calling capabilities under realistic API complexity by introducing WildAGTEval, a benchmark with 60 complexity scenarios and approximately 32K test configurations, finding that irrelevant information complexity reduced strong LLMs' performance by 27.3% and that LLMs sometimes distorted user intent to claim task completion.
We introduce WildAGTEval, a benchmark designed to evaluate large language model (LLM) agents' function-calling capabilities under realistic API complexity. Unlike prior work that assumes an idealized API system and disregards real-world factors such as noisy API outputs, WildAGTEval accounts for two dimensions of real-world complexity: 1. API specification, which includes detailed documentation and usage constraints, and 2. API execution, which captures runtime challenges. Consequently, WildAGTEval offers (i) an API system encompassing 60 distinct complexity scenarios that can be composed into approximately 32K test configurations, and (ii) user-agent interactions for evaluating LLM agents on these scenarios. Using WildAGTEval, we systematically assess several advanced LLMs and observe that most scenarios are challenging, with irrelevant information complexity posing the greatest difficulty and reducing the performance of strong LLMs by 27.3%. Furthermore, our qualitative analysis reveals that LLMs occasionally distort user intent merely to claim task completion, critically affecting user satisfaction.