robustness ↔ hallucination

Relations (1)

related 1.00 — strongly supporting 1 fact

These concepts are related as they are both key metrics used to evaluate the reliability and performance of large language models, as identified in the study by Zhang et al. (2023) [1].

Facts (1)

Sources

Building Trustworthy NeuroSymbolic AI Systems - arXiv arxiv.org arXiv 1 fact

claimZhang et al. (2023) identified reliability in LLMs by examining tendencies regarding hallucination, truthfulness, factuality, honesty, calibration, robustness, and interpretability.