AI in Mental Health: Emotional and Sentiment Analysis of Large Language Models' Responses to Depression, Anxiety, and Stress Queries

Arya VarastehNezhad; Hamed Farbeh; MohammadHossein LotfiNia; Reza Tavasoli; Soroush Elyasi

REVIEW 3 major objections 5 minor 69 references

AI in Mental Health: Emotional and Sentiment Analysis of Large Language Models' Responses to Depression, Anxiety, and Stress Queries

T0 review · 3 major / 5 minor · reviewed 2026-08-05 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read Anxiety prompts pushed eight large language models to near-ceiling fear scores (0.974), depression prompts to elevated sadness (0.686), and stress prompts to the most optimistic responses (0.755), with model identity shifting the emotional

desk verdict Systematic emotional-tone map with model-level differences that likely hold up, but the condition-level headline numbers are probably an artifact of tweet-trained classifiers keying on prompt words like 'anxiety'. read the letter →

arxiv 2508.11285 v1 pith:3XAVTTXR submitted 2025-08-15 cs.CL

Arya VarastehNezhad , Reza Tavasoli , Soroush Elyasi , MohammadHossein LotfiNia , Hamed Farbeh This is my paper

classification cs.CL

keywords largelanguagemodelsmentalhealthemotionanalysissentimentdepressionanxietystresshuman-AIinteraction

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper tries to establish that the emotional tone of large language model responses to mental-health questions is not uniform: the model asked and the condition named in the query systematically change the measured sentiment and emotion, while demographic framing changes it only marginally. Across 2,880 answers from eight systems, anxiety questions produced fear scores near 0.974, depression questions produced sadness near 0.686 and the most negative sentiment, and stress questions produced the most optimistic responses at 0.755. The authors argue that each model carries a distinct emotional signature, with Mixtral expressing the most disapproval, annoyance, and sadness and Llama the most optimism and joy. A sympathetic reader would care because users seeking mental-health guidance are likely receiving different emotional registers depending on which system they happen to ask.

What carries the argument

The measuring instrument is a stack of text classifiers trained on social-media data: a RoBERTa sentiment model built through TweetEval [11], a RoBERTa-large multilabel emotion model from SuperTweetEval [6], and models trained on the GoEmotions dataset [17]. These convert each LLM answer into a vector of sentiment and emotion intensities. The experimental design multiplies 20 pragmatic questions by 6 demographic framings by 3 conditions by 8 models, producing 2,880 answers whose score vectors are compared with Mann–Whitney U tests.

What would settle it

Re-annotation check: have clinicians or trained raters score a sample of the 2,880 responses, and separately rerun the same classifiers on anxiety prompts that avoid the word 'anxiety' (for example, 'racing heart, dread, unable to relax'). If human ratings diverge from the automated scores, or the 0.974 fear score drops sharply when the trigger word is removed, the headline magnitudes are lexical artefacts; if the pattern survives both checks, the claim of condition-adaptive emotional expression is supported.

Watch

Extended reading notes

Core claim

The authors' central discovery is that LLMs are not emotionally neutral answer engines in mental-health contexts. Using probabilistic text classifiers, they assign each generated answer scores for negative, neutral, and positive sentiment and for a set of emotions; on those scores, the model identity is a strong determinant of emotional expression, and the named condition is an even stronger one. Anxiety prompts drew fear responses near the classifier ceiling, at 0.974, while suppressing optimism and trust; depression prompts drew sadness at 0.686 and the highest negative sentiment; and stress prompts drew optimism at 0.755 with elevated joy and trust. The paper interprets this condition-dep

Load-bearing premise

The load-bearing premise is that classifiers trained on tweets score clinical-style LLM responses the same way they score social-media text, rather than over-weighting trigger words such as 'anxiety'; if that fails, the headline condition differences are partly lexical artefacts, though the model-vs-model comparisons on identical prompts may survive.

Editorial extensions

If this is right

Model choice changes the emotional register a user receives: Llama was the most optimistic and joyful system, Mixtral the most disapproving, annoyed, and sad, and the differences were statistically significant.
The condition named in the query drives emotional output: anxiety prompts produce near-ceiling fear, depression prompts sadness and negative sentiment, and stress prompts optimism, joy, and trust.
Demographic framing in the query has only marginal emotional effects; the profile with no demographic information showed the least positive and most negative sentiment, but the paper reads the overall demographic influence as minimal.
Organizations building mental-health applications should emotionally profile candidate models rather than choosing on technical metrics alone, matching the model's emotional signature to the application's goals.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

Because the classifiers were trained on tweets and never validated on clinical-style text, the headline condition differences may partly reflect the trigger word 'anxiety' rather than response tone; a paraphrase test that removes diagnostic labels would separate the two, and the paper's same-prompt model comparisons would likely survive such a test even if the magnitudes did not.
The paper's demographic conclusion is limited to how the query is framed in a single sentence; it does not test conversational context, self-disclosure, or intersectional identities, so a stronger bias conclusion would need those conditions.
If the emotional signatures are stable across paraphrases and over time, developers could calibrate mental-health interfaces to a desired emotional register; the study itself is a cross-sectional snapshot and does not establish that stability.
The authors note in Section 5 that the emotion-detection frameworks may miss the full nuance of therapeutic communication; paired human ratings would be the natural check on all of the automated magnitudes.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Short version: this is a useful empirical map, and the model-level result is probably real; the condition-level headline is not. The 2880-response grid (8 models x 3 conditions x 6 profiles x 20 questions) is a clean design, and the finding that model identity drives emotional tone far more than demographic framing is worth having. I'd trust the model comparisons because every model saw identical prompts; whatever bias the classifiers have largely cancels out across models. The demographic null result also looks believable.

The soft spot is the one the stress-test flags, and it's load-bearing for the abstract. The fear mean of 0.974 for anxiety prompts sits at the sigmoid ceiling, and every response is about a condition whose name is strongly fear-associated in the classifier's Twitter training data. The same logic applies to sadness/depression and optimism/stress. The authors use off-the-shelf tweet-trained models without any validation on LLM-generated clinical-style text. Their limitation paragraph says the frameworks 'may not capture the full nuance,' but that isn't the same as showing they discriminate real emotional content here. The condition-level magnitudes in the abstract are likely lexical echoes, not evidence of models 'mirroring' user emotion. The Section 5 interpretation goes past what the data support.

Secondary issues: thousands of Mann-Whitney tests without any multiple-comparison correction, no effect sizes or confidence intervals, and no data/code release. Copilot and Perplexity went through web interfaces while the rest used the OpenRouter API, which entangles channel with model identity for those two. All fixable, but right now the quantitative precision in Tables 1 and 2 is overstated.

I don't think this is a desk reject. The model-level emotional-signature differences are a genuine contribution that could survive a validated instrument, and the study fills a real gap. But it needs major revision: validate the classifiers on in-domain data (or use a validated clinical/psychological text analyzer), redo the stats with corrections and effect sizes, and release the response corpus. With those changes it could be a solid empirical reference. Without them, the headline is misleading.

For peer review: yes, send it out. The flaws are serious but standard, and the core question matters enough that a good referee cycle could turn this into a useful paper.

Referee Report

3 major / 5 minor

Summary. The paper reports an observational study of eight LLM-based systems (Claude, Copilot, Gemini, GPT-4o, GPT-4o mini, Llama, Mixtral, Perplexity) responding to 20 mental-health questions under three conditions (depression, anxiety, stress) and six demographic framings. The authors score 2,880 responses with Twitter-trained sentiment/emotion classifiers and GoEmotions-based models, then use Mann–Whitney U tests to compare emotion and sentiment scores across models, conditions, and demographics. The central claims are that each LLM has a distinct emotional signature, that the mental-health condition dramatically shapes emotional tone (anxiety fear 0.974, depression sadness 0.686, stress optimism 0.755), and that demographic framing has only marginal effects.

Significance. If the measurements are valid, the paper offers actionable evidence that model choice could materially affect the emotional tone of AI-provided mental-health information, and it proposes emotional profiling as a deployment criterion. The study design has genuine strengths: a relatively large corpus of 2,880 responses, eight systems queried with identical prompts within each condition, a clear question set, and statistical comparisons with reported p-values. However, the headline condition-level magnitudes rest on tweet-trained emotion classifiers that are never validated on LLM-generated clinical-style text. Because every anxiety prompt contains the word 'anxiety', every depression prompt contains 'depression', and every stress prompt contains 'stress', the near-ceiling fear score of 0.974 and the corresponding sadness/optimism contrasts are consistent with lexical triggering in the classifiers' Twitter training distributions rather than with the emotional content of the responses. This measurement-validity issue is load-bearing for the manuscript's most prominent claim, although the model-vs-model comparisons are less vulnerable because all models saw identical prompts.

major comments (3)

[§3.6 and Table 2] The condition-level magnitudes in the abstract (fear 0.974, sadness 0.686, optimism 0.755) are computed with cardiffnlp/twitter-roberta-large-emotion-latest and the TweetEval sentiment model, both trained on social-media text. The paper provides no validation of these classifiers on the target domain: LLM-generated, clinical-style mental-health responses. Since the anxiety prompts all contain the token 'anxiety', and that token is strongly fear-associated in Twitter data, the near-ceiling fear mean is equally explained by lexical association as by the actual tone of the responses. The same confound applies to 'depression'→sadness and 'stress'→optimism. I ask the authors to provide per-condition example responses, an ablation that masks or removes the condition word, or in-domain human-annotated validation. Without this, the 'affective mirroring' interpretation in §5 is circular rather th
[§3.5 and Table 1] Copilot and Perplexity were queried through their official web interfaces while the other six models were queried through the OpenRouter API. This entangles model identity with interface/system-prompt/version differences. Consequently, model-level comparisons involving Copilot and Perplexity (e.g., 'Perplexity highest caring', 'Copilot lowest positive sentiment') cannot be cleanly attributed to the model itself. The authors should either access all systems through a common gateway or explicitly relegate these two systems to exploratory status.
[§3.7 and §4.2] The statistical procedure runs a very large number of Mann–Whitney U tests—focal group versus all others, for each emotion, sentiment, model, condition, and demographic category—with no multiple-comparison correction. Some reported p-values (e.g., p = 0.041 for stress anger; p = 0.049 for old-user positive sentiment) are unlikely to survive even basic FDR control. In addition, the 'focal versus all others' comparison uses overlapping observations, violating the independence assumption of the test. The authors should report effect sizes or confidence intervals and apply a correction such as Benjamini–Hochberg, or treat the demographic findings as exploratory.

minor comments (5)

[Table 1] The 'disgust' row reports an average of 0.21, while all eight model-specific values are 0.016–0.027. This appears to be a decimal/typing error and should be corrected.
[References] Reference [49] is labeled OpenAI sycophancy material, but the URL points to a WHO depression fact sheet. The reference and citation need correction.
[§3.6] The text says the TweetEmotion model identifies 'ten distinct emotions' but then lists eleven (anger, anticipation, disgust, fear, joy, love, optimism, pessimism, sadness, surprise, trust). Please reconcile the count.
[§4.1] Table 1 would be easier to interpret with standard deviations or confidence intervals, particularly for the small differences among models (e.g., approval, trust, caring), which are otherwise hard to distinguish from noise.
[§5] The discussion states that models 'implicitly validate users’ experiences' through emotional congruence. This causal/experiential language goes beyond what the observational data can support; the data only show classifier scores, not user perception or validation.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the emotion/sentiment measurements rest on externally trained classifiers and the authors' self-citations are background only.

full rationale

The paper's central claims are empirical measurements: prompts are constructed, LLMs generate text, and externally published Twitter-trained classifiers (cardiffnlp/twitter-roberta-large-emotion-latest, TweetEval sentiment model) assign sentiment/emotion scores. No parameter is fitted to the outcome, no equation defines the condition-level result in terms of the instruments, and no load-bearing step is justified solely by a self-citation. The condition-level findings (e.g., anxiety prompts producing fear 0.974) could be confounded by the lexical content of the prompts and the classifiers' Twitter training distribution, but that is a construct-validity threat, not a circular reduction: the classifiers are independent, externally trained models, and their outputs are falsifiable. The paper itself acknowledges a related limitation in Section 5: 'the emotion detection frameworks used, while state-of-the-art, may not capture the full nuance of therapeutic communication.' This caveat weighs on validity, not on circularity. The self-citations ([19], [21], [55], [58], [59]) appear in related work and methodology as background and do not carry the derivation. The demographic/channel confounds (Copilot and Perplexity via web interfaces) are also methodological risks, not circularities. Therefore the derivation chain is self-contained given its stated instruments and assumptions.

Assumptions & free parameters 0 free parameters · 6 assumptions · 0 invented entities

The authors fit no parameters; all numbers are classifier outputs, LLM responses, or test statistics. The effective free parameters live in the externally fitted tweet classifiers, which are treated as a fixed measurement instrument without in-domain validation (listed under axioms). No new entities are postulated.

assumptions (6)

domain assumption Tweet-trained classifiers measure true emotional tone in LLM clinical-style text
Sections 3.6 and 4 treat scores from twitter-roberta-large-emotion-latest and the TweetEval sentiment model as measurements of expressed emotion with no in-domain validation; the anxiety fear mean of 0.974 suggests keyword-driven saturation.
domain assumption Default API parameters generate typical, representative responses
Section 3.5 keeps default parameters for all models; varying temperature or system prompts could shift emotional tone, and no sensitivity analysis is given.
domain assumption One generation per prompt adequately characterizes each model
Section 3.5 submits each query once; generation stochasticity is unmeasured, so model-level signatures mix true differences with sampling noise.
domain assumption The 20 questions broadly represent mental health information needs
Section 3.4 states coverage of diagnosis, treatment, prognosis, and support, but there is no external validation that these are the questions users actually ask.
domain assumption Mann-Whitney U assumptions hold with each response an independent observation
Section 3.7 treats 2,880 responses as exchangeable, though responses share prompts, models, and conditions, and no multiple-comparison correction is applied across thousands of tests.
domain assumption Models accessed in early 2025 stand for the named models
Section 3.5 gives a vague collection window and mixed channels (OpenRouter API versus web interfaces); commercial models update frequently, so the snapshot is not pinned.

how reviews work

0 comments

Cite this review

Pith. "Pith review of AI in Mental Health: Emotional and Sentiment Analysis of Large Language Models' Responses to Depression, Anxiety, and Stress Queries." pith.science (2026). https://pith.science/paper/3XAVTTXR

@misc{pith2026250811285,
  author       = {Pith},
  title        = {Pith review of: AI in Mental Health: Emotional and Sentiment Analysis of Large Language Models' Responses to Depression, Anxiety, and Stress Queries},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/3XAVTTXR}},
  note         = {Machine review of arXiv:2508.11285}
}

read the original abstract

Depression, anxiety, and stress are widespread mental health concerns that increasingly drive individuals to seek information from Large Language Models (LLMs). This study investigates how eight LLMs (Claude Sonnet, Copilot, Gemini Pro, GPT-4o, GPT-4o mini, Llama, Mixtral, and Perplexity) reply to twenty pragmatic questions about depression, anxiety, and stress when those questions are framed for six user profiles (baseline, woman, man, young, old, and university student). The models generated 2,880 answers, which we scored for sentiment and emotions using state-of-the-art tools. Our analysis revealed that optimism, fear, and sadness dominated the emotional landscape across all outputs, with neutral sentiment maintaining consistently high values. Gratitude, joy, and trust appeared at moderate levels, while emotions such as anger, disgust, and love were rarely expressed. The choice of LLM significantly influenced emotional expression patterns. Mixtral exhibited the highest levels of negative emotions including disapproval, annoyance, and sadness, while Llama demonstrated the most optimistic and joyful responses. The type of mental health condition dramatically shaped emotional responses: anxiety prompts elicited extraordinarily high fear scores (0.974), depression prompts generated elevated sadness (0.686) and the highest negative sentiment, while stress-related queries produced the most optimistic responses (0.755) with elevated joy and trust. In contrast, demographic framing of queries produced only marginal variations in emotional tone. Statistical analyses confirmed significant model-specific and condition-specific differences, while demographic influences remained minimal. These findings highlight the critical importance of model selection in mental health applications, as each LLM exhibits a distinct emotional signature that could significantly impact user experience and outcomes.

Figures

Figures reproduced from arXiv: 2508.11285 by the authors.

**Figure 1.** Overview of the research methodology, illustrating the selection of LLMs, demographic categories, mental health [PITH_FULL_IMAGE:figures/full_fig_p006_1.png] view at source ↗

**Figure 2.** Average emotional intensity scores across responses generated by the eight LLMs, categorized into positive, neutral, [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗

**Figure 3.** Box plot showing the distribution and variability of emotion scores across all LLM responses, highlighting the wide [PITH_FULL_IMAGE:figures/full_fig_p010_3.png] view at source ↗

Figures from the paper (1 more)

**Figure 4.** Figure 4: Correlation heatmap illustrating relationships among emotional features across LLM responses, with clusters of [PITH_FULL_IMAGE:figures/full_fig_p013_4.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

69 extracted references · 62 canonical work pages

[1]

Francisca Adoma Acheampong, Chen Wenyu, and Henry Nunoo-Mensah. 2020. Text-based emotion detection: Advances, challenges, and opportunities. Engineering Reports 2, 7 (2020), e12189

work page 2020
[2]

Prottay Kumar Adhikary, Aseem Srivastava, Shivani Kumar, Salam Michael Singh, Puneet Manuja, Jini K Gopinath, Vijay Krishnan, Swati Kedia Gupta, Koushik Sinha Deb, and Tanmoy Chakraborty. 2024. Exploring the efficacy of large language models in summarizing mental health counseling sessions: benchmark study. JMIR Mental Health 11 (2024), e57306

work page 2024
[3]

Doaa Alamoudi, Ian Nabney, and Esther Crawley. 2024. Evaluating the effectiveness of the SleepTracker app for detecting anxiety-and depression-related sleep disturbances. Sensors 24, 3 (2024), 722

work page 2024
[4]

Fahad Alanezi. 2024. Assessing the effectiveness of ChatGPT in delivering mental health support: a qualitative study. Journal of multidisciplinary healthcare (2024), 461–471. , Vol. 1, No. 1, Article . Publication date: September 2025. 16 • VarastehNezhad et al

work page 2024
[5]

Ahmed M Alhuwaydi. 2024. Exploring the role of artificial intelligence in mental healthcare: current trends and future directions–a narrative review for a comprehensive insight. Risk Management and Healthcare Policy (2024), 1339–1348

work page 2024
[6]

Dimosthenis Antypas, Asahi Ushio, Francesco Barbieri, Leonardo Neves, Kiamehr Rezaee, Luis Espinosa-Anke, Jiaxin Pei, and Jose Camacho-Collados. 2023. Supertweeteval: A challenging, unified and heterogeneous benchmark for social media nlp research. arXiv preprint arXiv:2310.14757 (2023)

work page Pith review arXiv 2023
[7]

Kristen A Archbell and Robert J Coplan. 2022. Too anxious to talk: Social anxiety, academic communication, and students’ experiences in higher education. Journal of emotional and behavioral disorders 30, 4 (2022), 273–286

work page 2022
[8]

William Babonnaud, Estelle Delouche, and Mounir Lahlouh. 2024. The bias that lies beneath: Qualitative uncovering of stereotypes in large language models. Swedish Artificial Intelligence Society (2024), 195–203

work page 2024

Show all 69 references

[9]

Jean-Christophe Bélisle-Pipon. 2024. Why we need to be careful with LLMs in medicine. Frontiers in Medicine 11 (2024), 1495582

2024
[10]

Felix Busch, Lena Hoffmann, Christopher Rueger, Elon HC van Dijk, Rawen Kader, Esteban Ortiz-Prado, Marcus R Makowski, Luca Saba, Martin Hadamitzky, and Jakob Nikolas Kather. 2025. Current applications and challenges in large language models for patient care: a systematic revi...

2025
[11]

Jose Camacho-Collados, Kiamehr Rezaee, Talayeh Riahi, Asahi Ushio, Daniel Loureiro, Dimosthenis Antypas, Joanne Boisson, Luis Espinosa-Anke, Fangyu Liu, and Eugenio Martínez-Cámara. 2022. TweetNLP: Cutting-edge natural language processing for social media. arXiv preprint arXiv...

2022 arXiv
[12]

Rosalyn Denise Campbell and Linda A Long. 2014. Culture as a social determinant of mental and behavioral health: A look at culturally shaped beliefs and their impact on help-seeking behaviors and service use patterns of Black Americans with depression. Best Practices in Mental...

2014
[13]

Lea Canales and Patricio Martínez-Barco. 2014. Emotion detection from text: A survey. In Proceedings of the workshop on natural language processing in the 5th information systems research working days (JISIC) . 37–43

2014
[14]

Mirko Casu, Sergio Triscari, Sebastiano Battiato, Luca Guarnera, and Pasquale Caponnetto. 2024. AI chatbots for mental health: A scoping review of effectiveness, feasibility, and applications. Appl. Sci 14 (2024), 5889

2024
[15]

Jan Cegin, Jakub Simko, and Peter Brusilovsky. 2024. LLMs vs Established Text Augmentation Techniques for Classification: When do the Benefits Outweight the Costs? arXiv preprint arXiv:2408.16502 (2024)

2024 arXiv
[16]

Ananya Kumari Dayanand, Rohit Tanwar, and Shahina Anwarul. [n. d.]. Harnessing Emotion Detection in Healthcare: Techniques, Challenges, and Future Directions. Sentiment Analysis Unveiled ([n. d.]), 131–150
[17]

Dorottya Demszky, Dana Movshovitz-Attias, Jeongwoo Ko, Alan Cowen, Gaurav Nemade, and Sujith Ravi. 2020. GoEmotions: A dataset of fine-grained emotions. arXiv preprint arXiv:2005.00547 (2020)

2020 arXiv
[18]

Yucong Duan, Fuliang Tang, Kunguang Wu, Zhendong Guo, Shuaishuai Huang, Yingtian Mei, Yuxing Wang, Zeyu Yang, and Shiming Gong. 2024. The large language model (llm) bias evaluation (age bias). DIKWP Research Group International Standard Evaluation. DOI 10 (2024)

2024
[19]

Soroush Elyasi, Arya Varasteh Nezhad, and Fattaneh Taghiyareh. 2023. Exploring the Relationship Between Gameplay Log Data and Depression & Anxiety. In 2023 14th International Conference on Information and Knowledge Technology (IKT) . IEEE, 50–56

2023
[20]

Soroush Elyasi and Fattaneh Taghiyareh. 2023. MBTI-Based Personality Assessment through Introducing a Puzzle Game. In 2023 9th International Conference on Web Research (ICWR) . IEEE, 102–107

2023
[21]

Soroush Elyasi, Arya VarastehNezhad, and Fattaneh Taghiyareh. 2025. From Play to Prediction: Assessing Depression and Anxiety in Players Behavior with Machine Learning Models. International Journal of Serious Games 12, 1 (2025), 83–102

2025
[22]

Catherine K Ettman and Sandro Galea. 2023. The potential influence of AI on population mental health. JMIR Mental Health 10 (2023), e49936

2023
[23]

Faiza Farhat. 2024. ChatGPT as a complementary mental health resource: a boon or a bane. Annals of Biomedical Engineering 52, 5 (2024), 1111–1114

2024
[24]

Mirna Fawaz and Ali Samaha. 2021. E-learning: Depression, anxiety, and stress symptomatology among Lebanese university students during COVID-19 quarantine. Nursing Forum 56, 1 (2021), 52–57

2021
[25]

Mary Jane Friedrich. 2017. Depression is the leading cause of disability around the world. Jama 317, 15 (2017), 1517–1517. doi:10.1001/ jama.2017.3826

2017
[26]

Isabel O Gallegos, Ryan A Rossi, Joe Barrow, Md Mehrab Tanjim, Sungchul Kim, Franck Dernoncourt, Tong Yu, Ruiyi Zhang, and Nesreen K Ahmed. 2024. Bias and fairness in large language models: A survey. Computational Linguistics 50, 3 (2024), 1097–1179

2024
[27]

The Lancet Global Health. 2020. Mental health matters. The Lancet. Global Health 8, 11 (2020), e1352

2020
[28]

Yining Hua, Fenglin Liu, Kailai Yang, Zehan Li, Hongbin Na, Yi-han Sheu, Peilin Zhou, Lauren V Moran, Sophia Ananiadou, and Andrew Beam. 2024. Large language models in mental health care: a scoping review. arXiv preprint arXiv:2401.02984 (2024)

2024 arXiv
[29]

Kexin Huang, Jaan Altosaar, and Rajesh Ranganath. 2019. Clinicalbert: Modeling clinical notes and predicting hospital readmission. arXiv preprint arXiv:1904.05342 (2019)

2019 arXiv
[30]

Loukas Ilias, Spiros Mouzakitis, and Dimitris Askounis. 2023. Calibration of transformer-based models for identifying stress and depression in social media. IEEE Transactions on Computational Social Systems 11, 2 (2023), 1979–1990. , Vol. 1, No. 1, Article . Publication date: ...

2023
[31]

Nazish Imran, Aateqa Hashmi, and Ahad Imran. 2023. Chat-GPT: opportunities and challenges in child mental healthcare. Pakistan Journal of Medical Sciences 39, 4 (2023), 1191

2023
[32]

Yu Jin, Jiayi Liu, Pan Li, Baosen Wang, Yangxinyu Yan, Huilin Zhang, Chenhao Ni, Jing Wang, Yi Li, and Yajun Bu. 2025. The Applications of Large Language Models in Mental Health: Scoping Review. Journal of Medical Internet Research 27 (2025), e69284

2025
[33]

Nur Zafifa Kamarunzaman, Alice Shanthi, Z Md Nen, Norfarhana Zulkifli, and Nur Izzati Shamsuri. 2020. Mental health issues among university students and educators’ readiness to detect and help. International Journal of Academic Research in Business and Social Sciences 10, 9 (2...

2020
[34]

Anne-Kathrin Kleine, Eesha Kokje, Eva Lermer, and Susanne Gaube. 2023. Attitudes toward the adoption of 2 artificial intelligence– enabled mental health tools among prospective psychotherapists: Cross-sectional study. JMIR human factors 10 (2023), e46859

2023
[35]

Harnain Kour and Manoj K Gupta. 2022. An hybrid deep learning approach for depression prediction from user tweets using feature-rich CNN and bi-directional LSTM. Multimedia Tools and Applications 81, 17 (2022), 23649–23685

2022
[36]

Preethi Lahoti, Nicholas Blumm, Xiao Ma, Raghavendra Kotikalapudi, Sahitya Potluri, Qijun Tan, Hansa Srinivasan, Ben Packer, Ahmad Beirami, and Alex Beutel. 2023. Improving diversity of demographic representation in large language models via collective-critiques and self-votin...

2023 arXiv
[37]

Jinhyuk Lee, Wonjin Yoon, Sungdong Kim, Donghyeon Kim, Sunkyu Kim, Chan Ho So, and Jaewoo Kang. 2020. BioBERT: a pre-trained biomedical language representation model for biomedical text mining. Bioinformatics 36, 4 (2020), 1234–1240

2020
[38]

Shir Lissak, Nitay Calderon, Geva Shenkman, Yaakov Ophir, Eyal Fruchter, Anat Brunstein Klomek, and Roi Reichart. 2024. The colorful future of llms: Evaluating and improving llms as emotional supporters for queer youth. arXiv preprint arXiv:2402.11886 (2024)

2024 arXiv
[39]

Zilin Ma, Yiyang Mei, and Zhaoyuan Su. 2024. Understanding the benefits and challenges of using large language model-based conversational agents for mental well-being support. In AMIA Annual Symposium Proceedings, Vol. 2023. 1105

2024
[40]

Tesfa Mekonen, Gary CK Chan, Jason P Connor, Leanne Hides, and Janni Leung. 2021. Estimating the global treatment rates for depression: a systematic review and meta-analysis. Journal of Affective Disorders 295 (2021), 1234–1242

2021
[41]

Tamir Mendel, Nina Singh, Devin M Mann, Batia Wiesenfeld, and Oded Nov. 2025. Laypeople’s Use of and Attitudes Toward Large Language Models and Search Engines for Health Queries: Survey Study. Journal of Medical Internet Research 27 (2025), e64290

2025
[42]

Francesca Minerva and Alberto Giubilini. 2023. Is AI the future of mental healthcare? Topoi 42, 3 (2023), 809–817

2023
[43]

Khai-Nguyen Nguyen, Khai Le-Duc, Bach Phan Tat, Duy Le, Long Vo-Dang, and Truong-Son Hy. 2024. Sentiment Reasoning for Healthcare. arXiv preprint arXiv:2407.21054 (2024)

2024 arXiv
[44]

Madunil A Niriella, Pathum Premaratna, Mananjala Senanayake, Senerath Kodisinghe, Uditha Dassanayake, Anuradha Dassanayake, Dileepa S Ediriweera, and H Janaka de Silva. 2025. The reliability of freely accessible, baseline, general-purpose large language model generated patient...

2025
[45]

Nick Obradovich, Sahib S Khalsa, Waqas U Khan, Jina Suh, Roy H Perlis, Olusola Ajilore, and Martin P Paulus. 2024. Opportunities and risks of large language models in psychiatry. NPP—Digital Psychiatry and Neuroscience 2, 1 (2024), 8

2024
[46]

Julia Ohse, Bakir Hadžić, Parvez Mohammed, Nicolina Peperkorn, Janosch Fox, Joshua Krutzki, Alexander Lyko, Fan Mingyu, Xiaohu Zheng, and Matthias Rätsch. 2024. GPT-4 shows potential for identifying social anxiety from clinical interview data. Scientific Reports 14, 1 (2024), 1–12

2024
[47]

David B Olawade, Ojima Z Wada, Aderonke Odetayo, Aanuoluwapo Clement David-Olawade, Fiyinfoluwa Asaolu, and Judith Eberhardt
[48]

Jesutofunmi A Omiye, Haiwen Gui, Shawheen J Rezaei, James Zou, and Roxana Daneshjou. 2024. Large language models in medicine: the potentials and pitfalls: a narrative review. Annals of internal medicine 177, 2 (2024), 210–220

2024
[49]

OpenAI. 2025. Expanding on what we missed with sycophancy. https://www.who.int/news-room/fact-sheets/detail/depression. Accessed: 2025-07-07

2025
[50]

Clare E Palmer, Emily Marshall, Edward Millgate, Graham Warren, Michael P Ewbank, Elisa Cooper, Samantha Lawes, Malika Bouazzaoui, Alastair Smith, and Chris Hutchins-Joss. 2024. Combining AI and human support in mental health: A digital intervention with comparable effectivene...

2024
[51]

Yoonseo Park, Sewon Park, and Munjae Lee. 2024. Effectiveness of artificial intelligence in detecting and managing depressive disorders: Systematic literature review. Journal of Affective Disorders (2024)

2024
[52]

Natalia K Rojas, Sam Martin, Mario Cortina-Borja, Roz Shafran, Lana Fox-Smith, Terence Stephenson, Brian CF Ching, Anaïs d’Oelsnitz, Tom Norris, and Yue Xu. 2025. Health and Experiences During the COVID-19 Pandemic Among Children and Young People: Analysis of Free-Text Respons...

2025
[53]

Abel Salinas, Parth Shah, Yuzhong Huang, Robert McCormack, and Fred Morstatter. 2023. The unequal opportunities of large language models: Examining demographic biases in job recommendations by chatgpt and llama. In Proceedings of the 3rd ACM Conference on Equity and Access in ...

2023
[54]

Sarek A Shen, Carlos A Perez-Heydrich, Deborah X Xie, and Jason C Nellis. 2024. ChatGPT vs. web search for patient questions: what does ChatGPT do better? European Archives of Oto-Rhino-Laryngology 281, 6 (2024), 3219–3225

2024
[55]

Reza Tavasoli, Arya VarastehNezhad, Mostafa Masumi, and Fattaneh Taghiyareh. 2025. Analyzing the Mathematical Proficiency of Large Language Models in Computer Science Graduate Admission Tests. In 2025 29th International Computer Conference, Computer Society of Iran (CSICC) . IEEE, 1–5

2025
[56]

Vankayala Tejaswini, Korra Sathya Babu, and Bibhudatta Sahoo. 2024. Depression detection from social media text analysis using natural language processing techniques and hybrid deep learning model. ACM Transactions on Asian and Low-Resource Language Information Processing 23, ...

2024
[57]

Arun James Thirunavukarasu, Darren Shu Jeng Ting, Kabilan Elangovan, Laura Gutierrez, Ting Fang Tan, and Daniel Shu Wei Ting
[58]

Arya VarastehNezhad, Reza Tavasoli, Mostafa Masumi, Seyed Soroush Majd, and Mehrnoush Shamsfard. 2024. Evaluating LLMs in Persian News Summarization. In 2024 15th International Conference on Information and Knowledge Technology (IKT) . IEEE, 195–201

2024
[59]

Arya VarastehNezhad, Reza Tavasoli, Mostafa Masumi, and Fattaneh Taghiyareh. 2024. LLM Performance Assessment in Computer Science Graduate Entrance Exams. In 2024 11th International Symposium on Telecommunications (IST) . IEEE, 232–237

2024
[60]

Gleb D Vzorinab, Alexey M Bukinichac, Anna V Sedykha, Irina I Vetrovab, and Elena A Sergienkob. 2024. The Emotional Intelligence of the GPT-4 Large Language Model. Psychology in Russia: State of the art 17, 2 (2024), 85–99

2024
[61]

World Health Organization. 2023. Depressive disorder (depression). https://www.who.int/news-room/fact-sheets/detail/depression

2023
[62]

World Health Organization. 2023. Anxiety disorders. https://www.who.int/news-room/fact-sheets/detail/anxiety-disorders. Accessed: 2025-02-02

2023
[63]

World Health Organization. 2024. Anxiety disorders. https://www.nimh.nih.gov/health/topics/depression. Accessed: 2025-02-03

2024
[64]

Xuhai Xu, Bingsheng Yao, Yuanzhe Dong, Hong Yu, James A Hendler, Anind K Dey, and Dakuo Wang. 2023. Leveraging large language models for mental health prediction via online text data. (2023)

2023
[65]

Tianlin Zhang, Annika M Schoene, Shaoxiong Ji, and Sophia Ananiadou. 2022. Natural language processing applied to mental illness detection: a narrative review. NPJ digital medicine 5, 1 (2022), 46

2022
[66]

Mi Zhou, Xiaomei Song, Qin Hu, and Youbin Zhou. 2025. Evaluating ChatGPT-4o’s Web-Enhanced Responses in Patient Education: Ankle Stabilization Surgery as a Case Study. (2025)

2025
[67]

Andrea Zucchetti, Gabriele Nibbio, Luca Altieri, Lorenzo Bertorni, Irene Calzavara-Pinton, Elena Invernizzi, Nicola Necchini, Caterina Cerati, Laura Poddighe, and Viola Bulgari. 2024. Artificial intelligence applications in mental health: The state of the art. Italian Journal ...

2024
[2023]

Nature medicine 29, 8 (2023), 1930–1940

Large language models in medicine. Nature medicine 29, 8 (2023), 1930–1940

2023
[2024]

Journal of medicine, surgery, and public health (2024), 100099

Enhancing mental health with Artificial Intelligence: Current trends and future prospects. Journal of medicine, surgery, and public health (2024), 100099

2024

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.

[1] [1]

Francisca Adoma Acheampong, Chen Wenyu, and Henry Nunoo-Mensah. 2020. Text-based emotion detection: Advances, challenges, and opportunities. Engineering Reports 2, 7 (2020), e12189

work page 2020

[2] [2]

Prottay Kumar Adhikary, Aseem Srivastava, Shivani Kumar, Salam Michael Singh, Puneet Manuja, Jini K Gopinath, Vijay Krishnan, Swati Kedia Gupta, Koushik Sinha Deb, and Tanmoy Chakraborty. 2024. Exploring the efficacy of large language models in summarizing mental health counseling sessions: benchmark study. JMIR Mental Health 11 (2024), e57306

work page 2024

[3] [3]

Doaa Alamoudi, Ian Nabney, and Esther Crawley. 2024. Evaluating the effectiveness of the SleepTracker app for detecting anxiety-and depression-related sleep disturbances. Sensors 24, 3 (2024), 722

work page 2024

[4] [4]

Fahad Alanezi. 2024. Assessing the effectiveness of ChatGPT in delivering mental health support: a qualitative study. Journal of multidisciplinary healthcare (2024), 461–471. , Vol. 1, No. 1, Article . Publication date: September 2025. 16 • VarastehNezhad et al

work page 2024

[5] [5]

Ahmed M Alhuwaydi. 2024. Exploring the role of artificial intelligence in mental healthcare: current trends and future directions–a narrative review for a comprehensive insight. Risk Management and Healthcare Policy (2024), 1339–1348

work page 2024

[6] [6]

Dimosthenis Antypas, Asahi Ushio, Francesco Barbieri, Leonardo Neves, Kiamehr Rezaee, Luis Espinosa-Anke, Jiaxin Pei, and Jose Camacho-Collados. 2023. Supertweeteval: A challenging, unified and heterogeneous benchmark for social media nlp research. arXiv preprint arXiv:2310.14757 (2023)

work page Pith review arXiv 2023

[7] [7]

Kristen A Archbell and Robert J Coplan. 2022. Too anxious to talk: Social anxiety, academic communication, and students’ experiences in higher education. Journal of emotional and behavioral disorders 30, 4 (2022), 273–286

work page 2022

[8] [8]

William Babonnaud, Estelle Delouche, and Mounir Lahlouh. 2024. The bias that lies beneath: Qualitative uncovering of stereotypes in large language models. Swedish Artificial Intelligence Society (2024), 195–203

work page 2024