Publications by authors named "Riccardo Loconte"
Article Synopsis
- The AI research community has relied on ad-hoc benchmarks to evaluate the capabilities of Large Language Models (LLMs), revealing challenges in replicating human-like intelligence, particularly in tasks linked to prefrontal lobe functions.
- In testing the GPT-3.5, GPT-4, Claude2, and Llama2 models using neuropsychological assessments, GPT-3.5 displayed inconsistent results with weaknesses in planning and social cognition, while GPT-4 performed relatively well across most tasks, except for planning.
- Overall, these findings suggest that current LLMs exhibit emergent abilities but do not fully mimic human cognitive functioning, indicating the potential of existing neuropsychological tests to effectively assess their performance.
Human accuracy in detecting deception with intuitive judgments has been proven to not go above the chance level. Therefore, several automatized verbal lie detection techniques employing Machine Learning and Transformer models have been developed to reach higher levels of accuracy. This study is the first to explore the performance of a Large Language Model, FLAN-T5 (small and base sizes), in a lie-detection classification task in three English-language datasets encompassing personal opinions, autobiographical memories, and future intentions.
View Article and Find Full Text PDFThe accurate recognition of others' facial expressions is a core skill for social interactions. The left Dorsolateral Prefrontal Cortex (L-DLPFC) represents a key node in the network for facial emotion recognition. However, its specific role is still under debate.
View Article and Find Full Text PDF