LifeAgentBench: Benchmarking LLMs for Long-Horizon, Cross-Dimensional Lifestyle Health Reasoning
Ye Tian*, Zihao Wang*, Onat Gungor, Xiaoran Fan, Tajana Rosing
Conference on Empirical Methods in Natural Language Processing (EMNLP), main conference, 2026 (accepted).
Project Website · Project Code · Paper PDF
Recommended citation: Tian, Y., Wang, Z., Gungor, O., Fan, X., & Rosing, T. (2026). LifeAgentBench: Benchmarking LLMs for Long-Horizon, Cross-Dimensional Lifestyle Health Reasoning. Conference on Empirical Methods in Natural Language Processing (EMNLP), main conference. Accepted. Tian and Wang contributed equally. https://arxiv.org/abs/2601.13880
LifeAgentBench contains 22,573 questions spanning single-domain retrieval to long-horizon, cross-dimensional reasoning over structured lifestyle health records. The latest benchmark study evaluates 13 LLMs. LifeAgent is a training-free, tool-augmented baseline using question decomposition, iterative retrieval, and deterministic computation.
*Co-first authors.