Pith. sign in

REVIEW 4 major objections 7 minor 4 cited by

Synthetic Data Generation with LLM for Improved Depression Prediction

T0 review · 4 major / 7 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read LLM-generated interview synopses improve depression severity prediction, the paper claims, with a combined real-plus-synthetic training set producing the lowest error.

desk verdict The synthetic-data pipeline is plausible but the evaluation is confounded by dataset size and shared generator style; still worth a referee. read the letter →

arxiv 2411.17672 v1 pith:HSKD4CWM submitted 2024-11-26 cs.LG

Andrea Kang , Jun Yu Chen , Zoe Lee-Youngzie , Shuhao Fu This is my paper · ORCID
classification cs.LG
keywords syntheticdatalargelanguagemodelsdepressionpredictionPHQ-8augmentationchain-of-thoughtpromptingBERTclinicalinterviews
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 proposes a two-step pipeline in which an open-source large language model, Llama 3.2, distills therapist-interview transcripts into short synopses with sentiment analyses and then rewrites those synopses at new, randomly sampled PHQ-8 depression severity scores. The claim is that adding these synthetic synopses to the real training data improves a BERT model's ability to predict depression severity, with the combined model reaching the lowest errors in the study (RMSE 4.64, MAE 3.66) and beating larger or purpose-built baselines trained only on real data. If true, the pipeline offers a way to enlarge and rebalance small, sensitive mental-health datasets without distributing raw patient transcripts. The authors also report that synthetic points sit farther from real transcripts in embedding space than real points sit from each other, which they read as evidence of privacy protection.

What carries the argument

The load-bearing mechanism is a chain-of-thought prompting pipeline built on Llama 3.2. In the first link, the raw interview transcript is compressed into a synopsis and a sentiment analysis; in the second, the model writes a fresh storyline and sentiment that matches a randomly sampled PHQ-8 score (0 to 24), using the original summary only as a structural reference. This 'distill then rewrite' design simultaneously solves two problems: the short synopsis fits BERT's 512-token limit (average 184 words versus 1389 for raw transcripts), and the forced rewrite at a new severity level keeps the synthetic text from being a near-copy of any real transcript. The assigned PHQ-8 score is then used as the training label for that synthetic sample.

What would settle it

Ask clinicians, blind to the assigned score, to rate the depression severity of each of the 309 synthetic synopses; if their ratings do not track the PHQ-8 scores written into the prompts, then the synthetic labels are unreliable and the reported test gains could reflect the LLM's stereotypes rather than clinical signal. A second decisive check would be to train the same BERT model on synthetic data whose PHQ-8 labels have been randomly permuted; if performance stays the same, the model is not using the severity signal at all.

Watch

Extended reading notes

Core claim

The paper's central claim is that synthetic data produced by its chain-of-thought prompting pipeline is faithful enough to real clinical interviews to serve as effective training data, and distant enough to preserve privacy, so that augmenting the real DAIC-WOZ training set with 309 synthetic synopses yields better depression severity prediction than training on real data alone. In the authors' experiments, the BERT model trained on real and synthetic synopses together achieved RMSE 4.64 and MAE 3.66, outperforming a BERT trained on real synopses (5.59/4.71), a BERT trained on synthetic data alone (4.80/4.06), and the Dual Encoder and Random Forest baselines trained on real transcripts. The synthetic data also rebalanced the training set's severity distribution, which had only 46 of 189 patients above the PHQ-8 cutoff of 10.

Load-bearing premise

The load-bearing premise is that a synthetic synopsis generated from a prompt containing a target PHQ-8 score actually exhibits that level of depression severity; the score is used as the training label with no clinician check that the text matches it.

Editorial extensions

If this is right

  • If the central claim is correct, synthetic synopses can substitute for raw transcripts when privacy constraints block access to real data; the model trained on synthetic data alone (RMSE 4.80) already outperformed the BERT trained on real synopses.
  • Combining synthetic and real data rebalances skewed severity distributions, which should reduce the risk of biased predictions for severely depressed patients underrepresented in real corpora.
  • The synopsis-plus-sentiment representation itself becomes a reusable, compact input format for downstream mental-health NLP tasks, since it preserves performance while staying within model token limits.
  • LLMs such as Llama 3.2 can act as scalable generators of labeled clinical narratives, lowering the cost of building larger depression-detection datasets.
  • The pipeline's privacy argument implies that versions of the real transcripts can be shared in summarized, rewritten form without the closest-neighbor leakage seen in the raw data.

Reading between the lines

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

  • The paper's privacy evidence rests on embedding distances, which measure dissimilarity, not re-identification risk; a stronger test would be a membership-inference attack, and a fair reading of this paper should treat the privacy guarantee as suggestive rather than proven.
  • Because the conditioning score is the only severity signal the LLM sees, the pipeline is only as clinically meaningful as Llama 3.2's understanding of PHQ-8 anchors; a direct test would compare human ratings of synthetic texts against the assigned score.
  • The same distill-and-rewrite pipeline could be applied to other self-report instruments (PHQ-9, GAD-7) or other sensitive clinical corpora; the paper does not test this, but nothing in its design is specific to depression.
  • The finding that synthetic-only training beats real-synopsis training hints that the real training set's imbalance was actively hurting the model; rebalancing by resampling real data rather than synthesizing new data would isolate how much of the gain is due to balance alone.
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.

Referee Report

4 major / 7 minor

Summary. The paper proposes a two-step chain-of-thought pipeline in which Llama 3.2 first converts DAIC-WOZ clinical interview transcripts into concise synopses and sentiment analyses, then generates new synthetic synopses/sentiments conditioned on a randomly sampled PHQ-8 score. A BERT regressor is trained on real synopses, synthetic synopses, or the combination, with the combined dataset reported to achieve the best test RMSE of 4.64 and MAE of 3.66, outperforming a zero-shot GPT-4o baseline, a Random Forest baseline, and a Dual Encoder baseline. The paper additionally reports embedding-distance-based privacy metrics and a PCA fidelity visualization. The central claim is that synthetic data augmentation improves depression severity prediction while preserving privacy.

Significance. If the results are validated, the pipeline would offer a practical, privacy-aware approach to augmenting small, imbalanced mental-health datasets with an open-source LLM. The paper makes its code available, uses an open-source model, and addresses a real clinical need. The conceptual idea of generating severity-conditioned summaries is appealing. However, the current experimental design does not yet establish that the reported improvements come from the synthetic generation quality rather than from dataset-size or label-distribution effects, nor that the synthetic labels carry clinically valid signal. The significance is therefore conditional on addressing these validity concerns.

major comments (4)
  1. [Utility, Table 1] The main comparison is confounded by dataset size and label rebalancing. The combined training set contains 416 samples (107 real + 309 synthetic) versus 107 real samples for the BERT 'Train' condition, and the synthetic data is explicitly designed to oversample high PHQ-8 scores. The reported improvement from RMSE 5.59 to 4.64 could be due simply to having more training data or to a more balanced label distribution, not to any property of the LLM-generated synopses. The authors should include an ablation that trains BERT on an equally sized real dataset (e.g., by oversampling real transcripts) and on a rebalanced-only real dataset, so that the effect of the synthetic generation itself can be isolated.
  2. [Chain of Thought Prompting, Appendix prompt] The synthetic labels are produced by construction rather than measured from the text. The second-stage prompt explicitly instructs Llama 3.2 to 'create a new synopsis ... that matches a depression/PHQ8 score of {PHQ8 Score}', and that same sampled score is then used as the training label for the resulting text. There is no human or clinical rating to verify that the generated text actually reflects the assigned severity. If the LLM produces stereotyped depression language (e.g., more negative emotion words) that does not correspond to realistic clinical presentation, the BERT model can learn those stylistic markers. This concern is amplified because the test set is also composed of Llama 3.2-generated synopses from real transcripts, so the low test error may reflect learning the LLM's text-generation style rather than transferable clinical signal. The paper needs to validate synthetic label fidelity, for example by having clinicians rate a sample of synthetic synopses against their assigned PHQ-8 scores, or by evaluating on raw transcripts rather than LLM-generated summaries.
  3. [Table 1, Results and Discussions] No uncertainty quantification is reported. Each entry in Table 1 is a single number for RMSE and MAE, with no standard deviation across random seeds, no confidence intervals, and no significance tests. Given the small training set and the stochasticity of BERT fine-tuning, a 0.03 RMSE difference (e.g., 4.67 for Dual Encoder versus 4.64 for the combined BERT) is likely within noise. The authors should run multiple seeds and report means with variances, and perform paired significance tests where appropriate, before claiming that the combined model outperforms the baselines.
  4. [Privacy, Table 2] The privacy evaluation does not support the claim that the synthetic data protects participant privacy. The minimum-embedding-distance metric only shows that synthetic synopses are farther from real synopses than real synopses are from each other in BERT embedding space. This is expected, since the synthetic synopses are deliberately written as new stories, and it does not rule out membership inference, attribute inference, or re-identification through other means. The abstract and conclusion make broad claims about 'safeguarding participant privacy' that go beyond what this metric demonstrates. The authors should either temper these claims or evaluate privacy with a more meaningful threat model, such as a nearest-neighbor re-identification attack against the actual synthetic data and transcripts.
minor comments (7)
  1. [Related Work] The phrase 'despressive disorders' is a typo and should read 'depressive disorders.'
  2. [Chain of Thought Prompting] The sentence ending 'by emphasizing style transfer through in learning' appears to be missing the word 'context'; it should read 'through in-context learning.'
  3. [Privacy] The sentence 'we quantitatively evaluate the privacy of our synthetic data by analyzing we quantitatively evaluate' contains a duplicated clause and should be rewritten.
  4. [Appendix] The prompt template for the second stage is incomplete: the placeholder for the depression-level description ends with '"One of "' and the actual mapping from PHQ-8 score ranges to text descriptions is not provided.
  5. [Experimental Setup] The paper does not specify how the test set is preprocessed: whether the test transcripts are also converted to synopses with the same Llama 3.2 model, and if so, this should be stated explicitly because it affects the interpretation of the test error.
  6. [Model Details] Details on the sampling distribution for the random PHQ-8 scores are missing; the paper says 'randomly generated PHQ-8 score (ranging from 0 to 24)' but does not say whether the sampling is uniform or stratified, which is relevant given the goal of rebalancing.
  7. [Table 1] The GPT-4o baseline is a zero-shot model evaluated on synopses, not fine-tuned on the training set; this should be explicitly labeled as a zero-shot baseline in the table caption or text to avoid implying an equivalent comparison.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the reported improvement is evaluated on held-out real DAIC-WOZ data, so the score-conditioned synthetic labels do not by construction determine the test result.

full rationale

The central claim is that augmenting training data with LLM-generated synthetic synopses improves PHQ-8 prediction. The synthetic generation step does condition on a randomly sampled PHQ-8 score: 'These outputs, combined with a randomly generated PHQ-8 score (ranging from 0 to 24) and its corresponding depression level description, are then used in a second prompt chain to create a synthetic synopsis and sentiment analysis.' Thus, for synthetic examples, the training label is an input to the generator rather than an independent measurement of the generated text. This is a genuine label-fidelity and weak-supervision limitation: the paper never clinically validates that the generated synopsis actually exhibits the requested severity, and the reported RMSE/MAE improvement could in principle reflect LLM stylistic markers rather than transferable clinical signal. However, this is a validity and external-transfer risk, not a circular derivation. The utility experiment is benchmarked on held-out real DAIC-WOZ test data: the paper reports 'We evaluated each model using two metrics: Root Mean Squared Error (RMSE) and Mean Absolute Error (MAE)' with Table 1 presenting Test RMSE and Test MAE against real PHQ-8 labels, using synopses generated from real test transcripts without score conditioning. No parameter is fitted to the test set, no prediction is renamed from a fitted input, and the authors do not rely on self-citations or imported uniqueness theorems. The fidelity (PCA overlap) and privacy (embedding-distance) analyses are empirical checks on the generated corpus, not equations that reduce to the model inputs. Therefore the central claim is self-contained against an external benchmark; the score-conditioned label construction is a correctness concern to be weighed separately, not circularity.

Assumptions & free parameters 3 free parameters · 5 assumptions · 0 invented entities

The paper introduces no new physical or theoretical entities. Its central result rests on hand-set generation and training choices, especially the assignment of synthetic PHQ-8 scores, and on untested assumptions about LLM fidelity and embedding-distance privacy.

free parameters (3)
  • Synthetic PHQ-8 score sampling procedure = three variations per training transcript; exact scores not listed
    The balancing effect and the combined data distribution depend on which PHQ-8 scores are assigned to each synthetic sample. The paper says scores are randomly sampled from 0 to 24, but 309 samples from 107 training transcripts implies an unstated non-uniform procedure that over-samples elevated scores.
  • Llama 3.2 decoding hyperparameters = max tokens 300-400; repetition penalty 1.175
    Generation diversity, coherence, and length control all depend on these hand-set decoding values; no sensitivity analysis is provided.
  • BERT training and early-stopping rule = learning rate 1e-5; up to 200 epochs; stop when MSE stops decreasing on evaluation dataset
    The stopping criterion is underspecified (no patience, no named split), and this choice affects the reported test RMSE for every model.
assumptions (5)
  • domain assumption DAIC-WOZ PHQ-8 scores are reliable ground-truth depression severity labels.
    Supervised training and evaluation treat the scores in the dataset as correct severity measures. This is standard practice for the benchmark but is not verified in the paper.
  • ad hoc to paper Llama 3.2-3B-Instruct can preserve clinically relevant content in synopses and can generate text matching an arbitrary assigned PHQ-8 score.
    This is the load-bearing premise of the whole pipeline. It is invoked in 'Chain of Thought Prompting' when the model is asked to create a synopsis that matches the requested depression level, and it is never validated against human ratings.
  • domain assumption Minimum BERT embedding distance is a meaningful privacy metric.
    The privacy conclusion (Table 2) assumes that larger nearest-neighbor distances in embedding space imply resistance to re-identification. No membership-inference or attacker model is tested.
  • domain assumption The reported test metrics were not used for model selection.
    The vague early-stopping description ('evaluation dataset') leaves open the possibility that test-set information influenced model selection; the paper does not state that early stopping used only the development split.
  • domain assumption Overlap of PCA projections of BERT embeddings indicates that synthetic data preserves the statistical structure of real data.
    Figure 4 is presented as fidelity evidence, but PCA overlap on embeddings is a weak proxy for distributional fidelity; it does not test whether depression-severity gradients are preserved.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Synthetic Data Generation with LLM for Improved Depression Prediction." pith.science (2026). https://pith.science/paper/HSKD4CWM

@misc{pith2026241117672,
  author       = {Pith},
  title        = {Pith review of: Synthetic Data Generation with LLM for Improved Depression Prediction},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/HSKD4CWM}},
  note         = {Machine review of arXiv:2411.17672}
}
read the original abstract

Automatic detection of depression is a rapidly growing field of research at the intersection of psychology and machine learning. However, with its exponential interest comes a growing concern for data privacy and scarcity due to the sensitivity of such a topic. In this paper, we propose a pipeline for Large Language Models (LLMs) to generate synthetic data to improve the performance of depression prediction models. Starting from unstructured, naturalistic text data from recorded transcripts of clinical interviews, we utilize an open-source LLM to generate synthetic data through chain-of-thought prompting. This pipeline involves two key steps: the first step is the generation of the synopsis and sentiment analysis based on the original transcript and depression score, while the second is the generation of the synthetic synopsis/sentiment analysis based on the summaries generated in the first step and a new depression score. Not only was the synthetic data satisfactory in terms of fidelity and privacy-preserving metrics, it also balanced the distribution of severity in the training dataset, thereby significantly enhancing the model's capability in predicting the intensity of the patient's depression. By leveraging LLMs to generate synthetic data that can be augmented to limited and imbalanced real-world datasets, we demonstrate a novel approach to addressing data scarcity and privacy concerns commonly faced in automatic depression detection, all while maintaining the statistical integrity of the original dataset. This approach offers a robust framework for future mental health research and applications.

Figures

Figures reproduced from arXiv: 2411.17672 by the authors.

Figure 1
Figure 1. A flowchart of the Chain-of-Thought pipeline. The LLM was able to capture the transcript’s key moments in both the [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Data distribution for different PHQ-8 scores in the original (left) and combined datasets (right). [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 4
Figure 4. PCA visualization of paragraph embeddings from [PITH_FULL_IMAGE:figures/full_fig_p006_4.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 4 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. OpenAlex reports about 5 citations worldwide. Full citation record

  1. Clinical Communication Processing with Models Trained on LLM-Generated Synthetic Data: A Structured Survey and Novel Application Case Studies

    cs.CL 2026-08 conditional novelty 6.0 of 10

    Synthetic clinical communication generated by LLMs can train clinical NLP models in thirteen case studies, but only one is tested on real patient text, leaving transfer to authentic communication unproven.

  2. Foundation Model Empowered Synesthesia of Machines (SoM): AI-native Intelligent Multi-Modal Sensing-Communication Integration

    eess.SP 2025-06 conditional novelty 4.0 of 10

    The paper proposes a systematic classification and two roadmaps for using foundation models (LLMs and wireless foundation models) to design Synesthesia of Machines systems for 6G, with preliminary case-study evidence ...

  3. Multimodal Large Language Models for Image, Text, and Speech Data Augmentation: A Survey

    cs.CV 2025-01 conditional novelty 4.0 of 10

    A literature review cataloging LLM-based augmentation methods across image, text, and speech, with a taxonomy of techniques, limitations, and suggested fixes.

  4. Unlocking the Potential of Large Language Models in the Nuclear Industry with Synthetic Data

    cs.CL 2025-06 conditional novelty 2.0 of 10

    A pipeline converts CANDU textbook chapters into synthetic QA pairs using LLMs, embedding clustering, and similarity metrics, with no downstream validation.

Reference graph

Works this paper leans on

45 extracted references · 33 canonical work pages · cited by 4 Pith papers

  1. [1]

    Abd Rahman, R.; Omar, K.; Noah, S. A. M.; Danuri, M. S. N. M.; and Al-Garadi, M. A. 2020. Application of machine learning methods in mental health detection: a systematic review. Ieee Access, 8: 183952--183964

    work page 2020

  2. [2]

    American Psychological Association . 2021. Worsening mental health crisis pressures psychologist workforce. Accessed: 2024-11-24

    work page 2021

  3. [3]

    A.; Dervovic, D.; Mahfouz, M.; Tillman, R

    Assefa, S. A.; Dervovic, D.; Mahfouz, M.; Tillman, R. E.; Reddy, P.; and Veloso, M. 2020. Generating synthetic data in finance: opportunities, challenges and pitfalls. In Proceedings of the First ACM International Conference on AI in Finance, 1--8

    work page 2020

  4. [4]

    Bao, E.; P \'e rez, A.; and Parapar, J. 2024. Explainable depression symptom detection in social media. Health Information Science and Systems, 12(1): 47

    work page 2024

  5. [5]

    Borkman, S.; Crespi, A.; Dhakad, S.; Ganguly, S.; Hogins, J.; Jhang, Y.-C.; Kamalzadeh, M.; Li, B.; Leal, S.; Parisi, P.; et al. 2021. Unity perception: generate synthetic data for computer vision. arXiv preprint arXiv:2107.04259

    arXiv 2021

  6. [6]

    Brown, T.; Mann, B.; Ryder, N.; Subbiah, M.; Kaplan, J. D.; Dhariwal, P.; Neelakantan, A.; Shyam, P.; Sastry, G.; Askell, A.; Agarwal, S.; Herbert-Voss, A.; Krueger, G.; Henighan, T.; Child, R.; Ramesh, A.; Ziegler, D.; Wu, J.; Winter, C.; Hesse, C.; Chen, M.; Sigler, E.; Litwin, M.; Gray, S.; Chess, B.; Clark, J.; Berner, C.; McCandlish, S.; Radford, A.;...

    2020

  7. [7]

    S.; Dhakal, S.; and Chiong, F

    Chiong, R.; Budhi, G. S.; Dhakal, S.; and Chiong, F. 2021. A textual-based featuring approach for depression detection using machine learning classifiers and social media texts. Computers in Biology and Medicine, 135: 104499

    work page 2021

  8. [8]

    Chodavadia, P.; Teo, I.; Poremski, D.; Fung, D. S. S.; and Finkelstein, E. A. 2023. Prevalence and economic burden of depression and anxiety symptoms among Singaporean adults: results from a 2022 web panel. BMC psychiatry, 23(1): 104

    work page 2023

Show all 45 references
  1. [9]

    Danner, M.; Hadzic, B.; Gerhardt, S.; Ludwig, S.; Uslu, I.; Shao, P.; Weber, T.; Shiban, Y.; and Ratsch, M. 2023. Advancing mental health diagnostics: GPT-based method for depression detection. In 2023 62nd Annual Conference of the Society of Instrument and Control Engineers (...

    2023

  2. [10]

    Dash, S.; Yale, A.; Guyon, I.; and Bennett, K. P. 2020. Medical time-series data generation using generative adversarial networks. In Artificial Intelligence in Medicine: 18th International Conference on Artificial Intelligence in Medicine, AIME 2020, Minneapolis, MN, USA, Aug...

    2020

  3. [11]

    Devlin, J. 2018. Bert: Pre-training of deep bidirectional transformers for language understanding. arXiv preprint arXiv:1810.04805

    2018 arXiv

  4. [12]

    Farruque, N.; Goebel, R.; Sivapalan, S.; and Za \" ane, O. R. 2024. Depression symptoms modelling from social media text: an LLM driven semi-supervised learning approach. Language Resources and Evaluation, 1--29

    2024

  5. [13]

    Ghanadian, H.; Nejadgholi, I.; and Al Osman, H. 2024. Socially Aware Synthetic Data Generation for Suicidal Ideation Detection Using Large Language Models. IEEE Access

    2024

  6. [14]

    Gong, Y.; and Poellabauer, C. 2017. Topic Modeling Based Multi-modal Depression Detection. In Proceedings of the 7th Annual Workshop on Audio/Visual Emotion Challenge, AVEC '17, 69–76. New York, NY, USA: Association for Computing Machinery. ISBN 9781450355025

    2017

  7. [15]

    Hadzic, B.; Mohammed, P.; Danner, M.; Ohse, J.; Zhang, Y.; Shiban, Y.; and R \"a tsch, M. 2024. Enhancing early depression detection with AI: a comparative use of NLP models. SICE journal of control, measurement, and system integration, 17(1): 135--143

    2024

  8. [16]

    K.; Kim, S.-H.; Jhon, M.; Yang, H.-J.; and Pant, S

    Iyortsuun, N. K.; Kim, S.-H.; Jhon, M.; Yang, H.-J.; and Pant, S. 2023. A Review of Machine Learning and Deep Learning Approaches on Mental Health Diagnosis. Healthcare, 11(3)

    2023

  9. [17]

    Kokosi, T.; and Harron, K. 2022. Synthetic data in medical research. BMJ medicine, 1(1)

    2022

  10. [18]

    W.; Spitzer, R

    Kroenke, K.; Strine, T. W.; Spitzer, R. L.; Williams, J. B.; Berry, J. T.; and Mokdad, A. H. 2009. The PHQ-8 as a measure of current depression in the general population. Journal of affective disorders, 114(1-3): 163--173

    2009

  11. [19]

    H.; You, S

    Kweon, S.; Kim, J.; Kim, J.; Im, S.; Cho, E.; Bae, S.; Oh, J.; Lee, G.; Moon, J. H.; You, S. C.; et al. 2023. Publicly Shareable Clinical Large Language Model Built on Synthetic Clinical Notes. arXiv preprint arXiv:2309.00237

    2023 arXiv

  12. [20]

    Lau, C.; Zhu, X.; and Chan, W.-Y. 2023. Automatic depression severity assessment with deep learning using parameter-efficient tuning. Frontiers in Psychiatry, 14: 1160291

    2023

  13. [21]

    W.; Taylor, N.; Nevado-Holgado, A.; Cipriani, A.; and Kormilitzin, A

    Lorge, I.; Joyce, D. W.; Taylor, N.; Nevado-Holgado, A.; Cipriani, A.; and Kormilitzin, A. 2024. Detecting the Clinical Features of Difficult-to-Treat Depression using Synthetic Data from Large Language Models. arXiv preprint arXiv:2402.07645

    2024 arXiv

  14. [22]

    Loshchilov, I.; and Hutter, F. 2017. Decoupled Weight Decay Regularization. In International Conference on Learning Representations

    2017

  15. [23]

    Man, K.; and Chahl, J. 2022. A review of synthetic image data and its use in computer vision. Journal of Imaging, 8(11): 310

    2022

  16. [24]

    F.; Murtaza, G.; Zafar, S.; and Bano, A

    Murtaza, H.; Ahmed, M.; Khan, N. F.; Murtaza, G.; Zafar, S.; and Bano, A. 2023. Synthetic data generation: State of the art in health care domain. Computer Science Review, 48: 100546

    2023

  17. [25]

    Muzammel, M.; Salam, H.; and Othmani, A. 2021. End-to-end multimodal clinical depression recognition using deep neural networks: A comparative analysis. Computer Methods and Programs in Biomedicine, 211: 106433

    2021

  18. [26]

    D.; Heinz, M

    Nemesure, M. D.; Heinz, M. V.; Huang, R.; and Jacobson, N. C. 2021. Predictive modeling of depression and anxiety using electronic health records and a novel machine learning approach with artificial intelligence. Scientific reports, 11(1): 1980

    2021

  19. [27]

    Ohse, J.; Had z i \'c , B.; Mohammed, P.; Peperkorn, N.; Danner, M.; Yorita, A.; Kubota, N.; R \"a tsch, M.; and Shiban, Y. 2024. Zero-Shot Strike: Testing the generalisation capabilities of out-of-the-box LLM models for depression detection. Computer Speech & Language, 88: 101663

    2024

  20. [28]

    OpenAI. 2023. OpenAI GPT-3.5. https://chatgpt.com/g/g-F00faAwkE-open-a-i-gpt-3-5

    2023

  21. [29]

    OpenAI. 2024. Hello GPT-4o. https://openai.com/index/hello-gpt-4o/. Accessed: 2024-11-10

    2024

  22. [30]

    H.; Buddhitha, P.; Orabi, M

    Orabi, A. H.; Buddhitha, P.; Orabi, M. H.; and Inkpen, D. 2018. Deep learning for depression detection of twitter users. In Proceedings of the fifth workshop on computational linguistics and clinical psychology: from keyboard to clinic, 88--97

    2018

  23. [31]

    Picardi, A.; Lega, I.; Tarsitani, L.; Caredda, M.; Matteucci, G.; Zerella, M.; Miglio, R.; Gigantesco, A.; Cerbo, M.; Gaddini, A.; et al. 2016. A randomised controlled trial of the effectiveness of a program for early detection and treatment of depression in primary care. Jour...

    2016

  24. [32]

    Reif, E.; Ippolito, D.; Yuan, A.; Coenen, A.; Callison-Burch, C.; and Wei, J. 2021. A recipe for arbitrary text style transfer with large language models. arXiv preprint arXiv:2109.03910

    2021 arXiv

  25. [33]

    H.; Schindler-Gmelch, L.; Berking, M.; and Eskofier, B

    Sadeghi, M.; Egger, B.; Agahi, R.; Richer, R.; Capito, K.; Rupp, L. H.; Schindler-Gmelch, L.; Berking, M.; and Eskofier, B. M. 2023. Exploring the capabilities of a language model-only approach for depression detection in text data. In 2023 IEEE EMBS International Conference o...

    2023

  26. [34]

    F.; Herrera, A

    Santomauro, D. F.; Herrera, A. M. M.; Shadid, J.; Zheng, P.; Ashbaugh, C.; Pigott, D. M.; Abbafati, C.; Adolph, C.; Amlag, J. O.; Aravkin, A. Y.; et al. 2021. Global prevalence and burden of depressive and anxiety disorders in 204 countries and territories in 2020 due to the C...

    2021

  27. [35]

    A.; Rogers, J

    Shin, H.-C.; Tenenholtz, N. A.; Rogers, J. K.; Schwarz, C. G.; Senjem, M. L.; Gunter, J. L.; Andriole, K. P.; and Michalski, M. 2018. Medical image synthesis for data augmentation and anonymization using generative adversarial networks. In Simulation and Synthesis in Medical I...

    2018

  28. [36]

    W.; Scales, N.; Tanwani, A

    Singhal, K.; Azizi, S.; Tu, T.; Mahdavi, S.; Wei, J.; Chung, H. W.; Scales, N.; Tanwani, A. K.; Cole-Lewis, H. J.; Pfohl, S. J.; Payne, P. A.; Seneviratne, M. G.; Gamble, P.; Kelly, C.; Scharli, N.; Chowdhery, A.; Mansfield, P. A.; y Arcas, B. A.; Webster, D. R.; Corrado, G. S...

    2022

  29. [37]

    R.; and Li, Y

    Squires, M.; Tao, X.; Elangovan, S.; Gururajan, R.; Zhou, X.; Acharya, U. R.; and Li, Y. 2023. Deep learning and machine learning in psychiatry: a survey of current progress in depression detection, diagnosis and treatment. Brain Informatics, 10(1): 10

    2023

  30. [38]

    Sun, B.; Zhang, Y.; He, J.; Yu, L.; Xu, Q.; Li, D.; and Wang, Z. 2017. A Random Forest Regression Method With Selected-Text Feature For Depression Assessment. In Proceedings of the 7th Annual Workshop on Audio/Visual Emotion Challenge, AVEC '17, 61–68. New York, NY, USA: Assoc...

    2017

  31. [39]

    Thieme, A.; Belgrave, D.; and Doherty, G. 2020. Machine learning in mental health: A systematic review of the HCI literature to support the development of effective and implementable ML systems. ACM Transactions on Computer-Human Interaction (TOCHI), 27(5): 1--53

    2020

  32. [40]

    Touvron, H.; Lavril, T.; Izacard, G.; Martinet, X.; Lachaux, M.-A.; Lacroix, T.; Rozi \`e re, B.; Goyal, N.; Hambro, E.; Azhar, F.; Rodriguez, A.; Joulin, A.; Grave, E.; and Lample, G. 2023. LLaMA: Open and Efficient Foundation Language Models. ArXiv, abs/2302.13971

    2023 arXiv

  33. [41]

    Trotzek, M.; Koitka, S.; and Friedrich, C. M. 2018. Utilizing neural networks and linguistic metadata for early detection of depression indications in text sequences. IEEE Transactions on Knowledge and Data Engineering, 32(3): 588--601

    2018

  34. [42]

    Wang, Y.; Inkpen, D.; and Gamaarachchige, P. K. 2024. Explainable depression detection using large language models on social media data. In Proceedings of the 9th Workshop on Computational Linguistics and Clinical Psychology (CLPsych 2024), 108--126

    2024

  35. [43]

    P.; Goyal, S.; Zhao, F.; Chintagunta, B.; and Ward, J

    Yuan, D.; Rastogi, E.; Naik, G.; Rajagopal, S. P.; Goyal, S.; Zhao, F.; Chintagunta, B.; and Ward, J. 2024. A continued pretrained llm approach for automatic medical note generation. arXiv preprint arXiv:2403.09057

    2024 arXiv

  36. [44]

    , " * write output.state after.block = add.period write newline

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint howpublished institution isbn journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all...

  37. [45]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.