{"doi":"10.1145/3643540","title":"Mental-LLM","abstract":"Advances in large language models (LLMs) have empowered a variety of applications. However, there is still a significant gap in research when it comes to understanding and enhancing the capabilities of LLMs in the field of mental health. In this work, we present a comprehensive evaluation of multiple LLMs on various mental health prediction tasks via online text data, including Alpaca, Alpaca-LoRA, FLAN-T5, GPT-3.5, and GPT-4. We conduct a broad range of experiments, covering zero-shot prompting, few-shot prompting, and instruction fine-tuning. The results indicate a promising yet limited performance of LLMs with zero-shot and few-shot prompt designs for mental health tasks. More importantly, our experiments show that instruction finetuning can significantly boost the performance of LLMs for all tasks simultaneously. Our best-finetuned models, Mental-Alpaca and Mental-FLAN-T5, outperform the best prompt design of GPT-3.5 (25 and 15 times bigger) by 10.9% on balanced accuracy and the best of GPT-4 (250 and 150 times bigger) by 4.8%. They further perform on par with the state-of-the-art task-specific language model. We also conduct an exploratory case study on LLMs' capability on mental health reasoning tasks, illustrating the promising capability of certain models such as GPT-4. We summarize our findings into a set of action guidelines for potential methods to enhance LLMs' capability for mental health tasks. Meanwhile, we also emphasize the important limitations before achieving deployability in real-world mental health settings, such as known racial and gender bias. We highlight the important ethical risks accompanying this line of research.","journal":"Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies","year":2024,"id":416272,"datarank":0.0,"base_score":0.0,"endowment":0.0,"self_citation_contribution":0.0,"citation_network_contribution":0.0,"self_endowment_contribution":0.0,"citer_contribution":0.0,"corpus_percentile":null,"corpus_rank":null,"citation_count":185,"citer_count":0,"citers_with_citation_signal":0,"citers_with_endowment":0,"datacite_reuse_total":0,"is_dataset":false,"is_dataset_confidence":0.9421,"is_data_producer":false,"deposit_databanks":null,"is_oa":true,"file_count":0,"downloads":0,"has_version_chain":false,"published_date":"2024-01-01","fair_score":null,"fair_percentile":null,"algorithm_id":"datarank_citation_only_1hop_v6","ranking_scope":"data_only","authors":[{"id":1159664,"name":"Bingsheng Yao","orcid":"0009-0004-8329-4610","position":1,"is_corresponding":false},{"id":1159665,"name":"Yuanzhe Dong","orcid":"0009-0006-2013-1157","position":2,"is_corresponding":false},{"id":1159666,"name":"Saadia Gabriel","orcid":"0009-0001-9353-951X","position":3,"is_corresponding":false},{"id":496030,"name":"Hong Yu","orcid":"0000-0001-9263-5035","position":4,"is_corresponding":false},{"id":3043,"name":"James Hendler","orcid":"0000-0003-3056-1960","position":5,"is_corresponding":false},{"id":550204,"name":"Marzyeh Ghassemi","orcid":"0000-0001-6349-7251","position":6,"is_corresponding":false},{"id":900658,"name":"Anind K. Dey","orcid":"0000-0002-3004-0770","position":7,"is_corresponding":false},{"id":1159667,"name":"Dakuo Wang","orcid":"0000-0001-9371-9441","position":8,"is_corresponding":false},{"id":1032345,"name":"Xuhai Xu","orcid":"0000-0001-5930-3899","position":0,"is_corresponding":true}],"reference_count":76,"raw_metadata":null,"created_at":"2026-07-19T01:56:20.964471Z","pmid":"39925940","pmcid":null,"fwci":null,"citation_percentile":null,"influential_citations":0,"oa_status":null,"license":null,"views":0,"total_file_size_bytes":0,"version_count":0,"fair_f":null,"fair_a":null,"fair_i":null,"fair_r":null,"fair_zscore":null,"fair_rationale":null,"fair_model":null,"fair_agent_version":null,"fair_fulltext_source":null,"fair_has_llm":null,"fair_computed_at":null,"clinical_trials":[],"software_tools":[],"db_accessions":[],"linked_datasets":[],"topics":[]}