Interpretable Symptom Vectors for Depression in a Large Language Model

Ajay Subramanian; Allison Constant; Camille Wang; Corey J. Keller; Fangyi Zhu; Ravish Gupta

REVIEW

Interpretable Symptom Vectors for Depression in a Large Language Model

Fangyi Zhu, Ajay Subramanian, Allison Constant, Camille Wang, Ravish Gupta, Corey J. Keller

Computation and Language · Submitted 1 Sep 2026 · 40 references

Pith's verdict: Plausible extension of semantic projection to LLM residual streams for depression symptoms, with clear limitations on label independence that the paper itself acknowledges.

T0 · AI review

How this review was made: An AI referee (deepseek/deepseek-v4-flash) read the paper on September 8, 2026 and answered a fixed list of questions about it. T0 is the starting level every paper gets. The levels describe how deeply Pith has checked the paper, never the authors.

T0 AI review An AI read the paper and answered a fixed list of questions.
T1 Audit Automatic checks that the references and files the paper relies on exist and match.
T2 Compute The paper's code or calculations re-run with the same results.
T3 Formal statement The claims written out in a formal math language a computer can read.
T4 Proof Those claims proved, and the proof checked by computer.

How Pith checks papers →

· 4 major objections · 6 minor · Ask about this paper · The compiled record · Read the full review →

Challenge this review Comment (0) Read on arXiv ↗

Cite ▾

arXiv ID2609.01832 (version 1)

Pith IDG3HZ2MC5, Pith's permanent record of this paper and its review

CitationAjay Subramanian, Allison Constant, Camille Wang, Corey J. Keller, Fangyi Zhu, Ravish Gupta. "Interpretable Symptom Vectors for Depression in a Large Language Model." arXiv:2609.01832 (2026). https://pith.science/paper/G3HZ2MC5

Cite this reviewBibTeX for Pith's review of this paper

Claim paper

Why Pith says this

Highlight any text to comment on it.

Editor's note

read the letter

This paper adapts semantic projection to LLM residual streams and shows that layer-21 activations of Gemma-3-27B-PT encode symptom categories (mood, somatic, suicidality) in a way that preserves clinician-annotated rank order on held-out text. That is a solid subfield contribution for clinical NLP and mechanistic interpretability, and it is executed carefully: the Gram-pseudoinverse decorrelation handles collinearity well, the pipeline is clearly described, and the limitations section is unusually candid.

What is actually new is mapping a multi-symptom clinical taxonomy onto decorrelated per-symptom axes in an LLM's internal activations without fine-tuning. The held-out rank-ordering in Table 4 is real evidence, and the AUC=0.789 on depression vs. happy text is a useful sanity check.

The stress-test concern lands: the held-out Naturalistic labels for Darkness Visible and Handbook of Depression were assigned by the same two psychiatrists using the same DSM-5 groupings that define the Symptom Vectors. The mood axis has no externally labeled held-out samples—ReDSM5 supplies only somatic and suicidality. This means the generalization claim is weaker than advertised; the result is better described as showing internal consistency plus a partially independent held-out check, not fully independent validation. The operating layer was also chosen by in-sample maximization of PERMANOVA pseudo-F on the same corpus used to build the vectors, which introduces some risk of overfitting to that layer. And the Depression Vector contrast confounds valence with genre and text length. These are specific, addressable weaknesses rather than fatal flaws—the authors acknowledge most of them explicitly.

Who this is for: researchers working on clinical NLP, mechanistic interpretability of sentiment/emotion in LLMs, or applications of semantic projection. A reader in those areas will find the methodology adaptable and the limitations instructive. It deserves a serious referee, and with better characterization of the label-dependence issue (e.g., leave-one-annotator-out or ReDSM5-only analysis) it would be a clean paper.

I would bring it to reading group and cite it if I worked in this area.

Referee report

Significance

If the held-out effects survive the additional controls requested below, this would be a genuinely useful result for clinical NLP and mechanistic interpretability: a no-fine-tuning readout of a clinically motivated symptom taxonomy directly from residual stream activations, with an explicit distinction between in-sample consistency and held-out generalization. The methodological strengths are real: the PERMDISP eligibility gate prevents misreading dispersion differences as centroid shifts; the convergence of three raw distance metrics on layer 21 is a nontrivial robustness observation; the Core Clinical results are honestly labeled as in-sample checks; the Gram-pseudoinverse projection and the two-stage valence-gate design are clearly specified; code and part of the corpora are released; and the Discussion explicitly refuses to equate annotation agreement with clinical validity. The load-bearing weaknesses are the label independence of the held-out mood axis, the absence of matched neutral and negative controls for the Depression Vector, the in-sample selection of the operating layer, and source-level clustering in the statistical tests.

Major objections (4)

[§2.1.1, Table 4] The held-out mood-axis result is not supported by externally labeled data. Table 1 shows that ReDSM5 contributes no mood passages, so all 141 held-out mood examples come from Darkness Visible and the Handbook of Depression, which were annotated by the same two psychiatrist authors using the same DSM-5 A1–A9 item groupings that define the Symptom Vector axes (§2.1.1). The mood axis is therefore the first axis claimed in the abstract and the one whose held-out rank ordering cannot distinguish “residual stream encodes clinician consensus” from “residual stream encodes this pair of annotators’ grouping choices.” The within-axis median ordering in Table 4 does hold as reported, but the mood result is also the weakest in effect size: the held-out mood-group median on the mood axis is negative (−0.08, versus +0.77 in-sample), and on the suicidality axis the non-matching mood and somatic groups score almost as high (0.85 and 0.91) as the matching group (1.40). Please report the rank ordering and separation statistics separately for passages with externally sourced labels (the ReDSM5 somatic and suicidality subsets) versus author-annotated book excerpts, and obtain either independent third-party annotations of the book excerpts or an externally labeled corpus with mood items; without this, the clinician-aligned generalization claim for the mood axis is not established.
[§2.4, §3.3] The Depression Vector test compares two corpora that differ in emotional valence but also in genre, source, and register: Naturalistic is book-length clinical and memoir prose plus Reddit posts, while HappyDB is short crowdsourced happy-moment sentences, and the absolute cosine differences are tiny (median 0.229 versus 0.218 on an overall 0.18–0.25 band). The authors themselves note that the two HappyDB outliers with the highest Depression Scores are neutral texts (“Their sore throat cleared up”; “They went shopping”), which shows the score fires on non-depressive content and strengthens the need for a matched control. Please add a matched neutral corpus and a non-depressive negative-valenced control (for example, anxiety-themed text) of comparable length and register to both Naturalistic and HappyDB, and re-report the AUC with those controls; as it stands, the “emotional valence gate” interpretation of AUC = 0.789 is confounded with genre and text-length differences.
[§2.2.3] The operating layer ℓ* = 21 is selected by maximizing PERMANOVA pseudo-F on the same 51 Core Clinical excerpts from which the Symptom Vectors are then built, and no independent portion of the data is used to confirm the layer choice. The agreement of three raw distance metrics at layer 21 is suggestive, but these metrics are computed on the same passages and are not independent tests. Please provide a stability analysis, for example by selecting the layer on random subsets of the Core Clinical corpus, or by reporting the held-out rank ordering and the Naturalistic-versus-HappyDB AUC across a range of layers around 21; this would show whether “layer 21” is a reproducible property of the model or an artifact of in-sample selection.
[§3.2, Fig. 3] The inferential statistics treat the 423 held-out passages as independent samples, but the passages are heavily clustered: 172 excerpts come from a single handbook and 57 from a single memoir, and the three annotation groups have different source compositions (mood is 100% book-sourced, suicidality is about 75% ReDSM5, somatic is about 62% ReDSM5). The Dunn post-hoc and Mann–Whitney tests therefore overstate significance, and the between-group median differences in Table 4 could partly reflect source identity rather than symptom content. Please report source-level medians and cluster-robust or source-aggregated tests, and re-check the Table 4 rank ordering under permutations that respect the source structure.

Minor objections (6)

[§2.3, Eq. (6)] The paper motivates the Moore–Penrose pseudoinverse by the near-singularity of the Gram matrix (condition number 14,611.6), but for a full-rank 3×3 matrix G+ equals G−1 unless a truncation tolerance is applied; please report the tolerance (for example, the rcond parameter) actually used, and if no truncation was applied, state explicitly that the projection uses the ordinary inverse and justify the “regularization” language.
[Data availability] The data availability statement says the rewritten Core Clinical and Positive Affect corpora are released but does not state whether the rewritten Naturalistic corpus (the held-out evaluation set) is available; please state its availability status or the copyright basis for withholding it, since this corpus carries the main generalization claims.
[§2.1.1, Table 1] The balancing procedure is described only in general terms, and a reader cannot tell from Table 1 why ReDSM5 contributes zero mood passages; because the 33 Darkness Visible and 108 Handbook mood excerpts already reach the 141 target, no ReDSM5 mood posts were needed, and the paper should say so explicitly to avoid the appearance of a data gap.
[§2.4, Eqs. (7)–(10)] The Depression centroid in Eq. (7) is the unweighted mean of the three Symptom Vectors, while the Positive Affect centroid in Eq. (8) and the scored excerpt centroid in Eq. (10) are token-weighted means; since the Depression Vector and the resulting scores inherit this asymmetry, the choice should be stated explicitly and its effect on the reported cosine values discussed.
[§2.1.1] The construction of the Positive Affect corpus is described only by genre examples and a final count of nine excerpts; since this corpus defines the positive endpoint of the Depression Vector (Eq. (8)), the selection procedure (how many candidate texts were considered, who chose them, and why nine) should be specified.
[§2.1.2] Because the same third-person rewrite prompt was applied to the corpora that define the Symptom Vectors and the Depression Vector and to the held-out Naturalistic corpus, the evaluation establishes generalization only for the rewritten third-person register; the Discussion lists this as limitation three, but the abstract and introduction should make explicit that the validated input channel requires the same standardization step.

How Pith checks papers → · Full audit of this paper →

The paper

The paper claims

A pre-trained large language model's internal activations encode clinician-aligned depression symptoms, recoverable without fine-tuning.

The reading

This paper claims that the internal residual stream of an unmodified large language model (Gemma-3-27B-PT) contains distinct, decorrelated signals for three groups of depression symptoms (mood, somatic, suicidality) that align with clinician judgments. The authors construct Symptom Vectors from clinical instrument descriptions at layer 21, where the three categories are maximally separable. When held-out naturalistic narratives are projected onto these vectors, the resulting coefficients preserve the clinician-annotated rank ordering of symptom categories. Additionally, a single Depression Vector derived from the same layer distinguishes depressive from non-depressive text with an AUC of 0.789. If correct, this provides a mechanistic, interpretable method for extracting symptom-level information from patient speech without requiring model fine-tuning or generated text.

What carries the argument

The central object is the set of three Symptom Vectors (mood, somatic, suicidality) formed from class centroids of residual stream activations at layer 21, along with a Depression Vector defined as their centroid minus a positive affect centroid. The argument is carried by the Gram-pseudoinverse decorrelation procedure, which resolves the near-collinearity (condition number ~14,600) of the clinical category centroids into stable, interpretable projection coefficients that isolate each symptom's unique contribution.

Extended reading notes

Core claim

The residual stream activations of Gemma-3-27B-PT at transformer layer 21 contain linear directions (Symptom Vectors) corresponding to mood, somatic, and suicidality symptom groups, constructed from six validated clinical instruments. These vectors, decorrelated via a Gram-pseudoinverse correction, yield projection coefficients on held-out naturalistic text that preserve the same rank ordering as independent clinician annotation: each symptom group attains the highest median coefficient on its own axis. Moreover, a Depression Vector defined as the difference between the depression centroid and a positive affect centroid separates held-out depressive narratives from happy-moment descriptions (AUC=0.789), acting as a valence gate for subsequent symptom projection.

Load-bearing premise

The held-out Naturalistic corpus was annotated by the same two psychiatrist authors using the same DSM-5 item groupings that define the Symptom Vector axes, and every excerpt was rewritten into third-person prose by another large language model; if the annotation definitions or the rewrite bias both axes and held-out labels in the same direction, the observed rank ordering overstates true generalization.

Editorial extensions

If this is right

No fine-tuning or generated text is needed: symptom-level signals are directly readable from a pre-trained LLM's internal activations.
The two-stage pipeline (valence gate via Depression Vector, then symptom projection) could be applied to patient speech for continuous, per-symptom monitoring in research settings.
Clinician-aligned symptom coefficients provide a more granular alternative to traditional severity sum scores, preserving dimensional profiles.
The methodology offers a template for extracting other clinical constructs from LLM internal representations using validated instrument descriptions.

Reading between the lines

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

The approach likely extends to other mental health conditions (e.g., anxiety, PTSD) if similar clinically validated symptom descriptions are available to construct basis vectors.
The strong collinearity of symptom centroids (condition number 14,600) may indicate a fundamental representational overlap in the model's internal space, reflecting the real-world comorbidity of depression symptoms; this is handled by the pseudoinverse but limits resolution.
The single-model, single-layer finding should be tested across architectures (LLaMA, GPT) and scales; failure to replicate would suggest the effect is model-specific rather than general to transformer LMs.

Original abstract

Patients with depression present with diverse symptom profiles, yet clinical practice routinely reduces this variation to a single severity score. Large language models (LLMs) can potentially capture various symptoms and their severity from patient speech. However, how depressive symptoms are represented inside LLMs remains poorly understood, limiting clinical trust. To examine whether internal model activations match clinician judgment, we analyzed the residual stream of Gemma-3-27B-PT using mechanistic interpretability techniques. Recording activations across symptom descriptions drawn from validated clinical instruments, we found that symptom groups geometrically separated the most at layer 21 across multiple distance metrics. Using Semantic Projection, we then projected held-out naturalistic text onto Symptom Vectors constructed from these instruments. The resulting per-symptom coefficients preserved clinician-annotated rank ordering across mood, somatic, and suicidality axes. Furthermore, a single depression vector in Layer 21 separates held-out depressive from non-depressive text (AUC = 0.789), which can be used as an emotional valence gate that restricts symptom projection to depressive speech. These results reveal a decorrelated, clinician-aligned symptom signal readable directly from internal activations, offering a mechanistic foundation for interpretable depression-assessment tools.

The compiled record

The paper's own sentences, arranged, kept beside the review and separate from it. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. Numbers in the tables are the authors', checked by code. No sentence in this record was written by a model.

Compiled by classifier-1.13.0 via TypeSafe API, question bank abstract-p1-20260923, 2026-09-23: 1 request, 6,921 input tokens.

What the abstract itself states, as probabilities: claims first 0.05 · states limitation 0.24 · human participants 0.51 · controlled comparison 0.35 · randomized assignment 0.02 · states quantitative result 0.99 · compares with prior results 0.23 · mentions code or data availability 0.03.

The abstract, sentence by sentence

arXiv listing abstract v1. Results and the main claim are underlined; show every role.

Patients with depression present with diverse symptom profiles, yet clinical practice routinely reduces this variation to a single severity score. Large language models (LLMs) can potentially capture various symptoms and their severity from patient speech. However, how depressive symptoms are represented inside LLMs remains poorly understood, limiting clinical trust. To examine whether internal model activations match clinician judgment, we analyzed the residual stream of Gemma-3-27B-PT using mechanistic interpretability techniques. Recording activations across symptom descriptions drawn from validated clinical instruments, we found that symptom groups geometrically separated the most at layer 21 across multiple distance metrics. Using Semantic Projection, we then projected held-out naturalistic text onto Symptom Vectors constructed from these instruments. The resulting per-symptom coefficients preserved clinician-annotated rank ordering across mood, somatic, and suicidality axes. Furthermore, a single depression vector in Layer 21 separates held-out depressive from non-depressive text (AUC = 0.789), which can be used as an emotional valence gate that restricts symptom projection to depressive speech. These results reveal a decorrelated, clinician-aligned symptom signal readable directly from internal activations, offering a mechanistic foundation for interpretable depression-assessment tools.

background contribution method result interpretation main claim

What the classifier reports for each sentence. The number is the probability it gave the label.
#	sentence	role	the authors' own work	states the main conclusion	strength
s0	Patients with depression present with diverse symptom profiles, yet clinical practice routinely reduces this variation to a single severity score.	background 0.99	0.05	0.02	asserted 1.00
s1	Large language models (LLMs) can potentially capture various symptoms and their severity from patient speech.	background 0.98	0.06	0.02	hedged possibility 1.00
s2	However, how depressive symptoms are represented inside LLMs remains poorly understood, limiting clinical trust.	gap 1.00	0.05	0.03	asserted 0.99
s3	To examine whether internal model activations match clinician judgment, we analyzed the residual stream of Gemma-3-27B-PT using mechanistic interpretability techniques.	method 0.75	0.93	0.02	asserted 0.79
s4	Recording activations across symptom descriptions drawn from validated clinical instruments, we found that symptom groups geometrically separated the most at layer 21 across multiple distance metrics.	result 0.99	0.97	0.09	asserted 1.00
s5	Using Semantic Projection, we then projected held-out naturalistic text onto Symptom Vectors constructed from these instruments.	method 0.99	0.94	0.02	asserted 0.93
s6	The resulting per-symptom coefficients preserved clinician-annotated rank ordering across mood, somatic, and suicidality axes.	result 1.00	0.95	0.14	asserted 1.00
s7	Furthermore, a single depression vector in Layer 21 separates held-out depressive from non-depressive text (AUC = 0.789), which can be used as an emotional valence gate that restricts symptom projection to depressive speech.	result 1.00	0.94	0.16	asserted 1.00
s8	These results reveal a decorrelated, clinician-aligned symptom signal readable directly from internal activations, offering a mechanistic foundation for interpretable depression-assessment tools.	interpretation 1.00	0.91	0.93	asserted 1.00

What kind of paper the classifier reports this is: empirical study 0.88 · evidence from observation or experiment 0.61.

What was examined, and what was not

sources read

arXiv listing abstract v1

not examined

full text
figures and tables
references and citing sentences
relations to other papers

levels run

P abstract

not run

S sections
T body sentences
C citations
R cross-paper relations

Every quoted sentence on this record is the authors'. Every label is from Pith's closed vocabulary, with the probability shown beside it. Every number in the measurements table is the authors', checked by code. Record as JSON

Figures

**Figure 1.** Analysis pipeline. A Residual stream activations are extracted from Gemma-3-27B-PT across all 63 hook points (ℓ = 0, . . . , 62) while prompting with symptom descriptions from three groups. B For each layer × distance-metric combination, distance matrices are computed and tested with PERMANOVA and PERMDISP; the operating layer ℓ ∗ is the layer maximizing pseudo-F subject to PERMDISP P > 0.05. C Symptom Vectors are constructed from the group centroids at ℓ ∗ , and naturalistic text samples are projected onto them to obtain symptom coefficients β. D The Depression Vector is defined as

**Figure 2.** Layer 21 maximizes symptom separability. PERMANOVA pseudo-F (y-axis) across all 63 hook points (x-axis) for each of the eight distance-metric × normalization combinations. Shaded regions mark layers where PERMDISP is significant (P < 0.05) and pseudo-F is therefore ineligible as a measure of centroid separation; the operating maximum (filled marker) is the maximum pseudo-F within each metric’s eligible region (P > 0.05), and the excluded maximum (open marker) is a higher but ineligible global maximum where shown. 11

**Figure 3.** Semantic projection yields clinician-aligned symptom coefficients. Symptom coefficients (βmood, βsomatic, βsuicidality) projected onto Symptom Vectors, grouped by clinician annotation and decorrelated. A–C Per-axis coefficient distributions (box plots with overlaid points) for the mood (A), somatic (B), and suicidality (C) axes, with the in-sample Core Clinical corpus on the top row and the held-out Naturalistic corpus on the bottom; boxes show the median and interquartile range, points show individual excerpts, and n is the number of excerpts per group. Brackets denote pairwise…

Figures reproduced from arXiv:2609.01832 by the authors.

References

Works cited (40)

[1]

Fried, E.I. & Nesse, R.M. Depression is not a consistent syndrome: an investigation of unique symptom patterns in the STAR*D study.J. Affect. Disord.172, 96–102; 10.1016/j.jad.2014.10.010 (2015)

View on Pith doi:10.1016/j.jad.2014.10.010 2015
[2]

Chevance, A. et al. Identifying outcomes for depression that matter to patients, informal caregivers, and health-care professionals: qualitative content analysis of a large international online survey.Lancet Psychiatry7, 692–702; 10.1016/S2215-0366(20)30191-7 (2020)

View on Pith doi:10.1016/s2215-0366(20)30191-7 2020
[3]

Yang, K. et al. MentaLLaMA: interpretable mental health analysis on social media with large language models. InProc. ACM Web Conference 2024 (WWW ’24)4489–4500; 10.1145/3589334.3648137 (ACM, 2024)

View on Pith arXiv 2024
[4]

Xu, X. et al. Mental-LLM: leveraging large language models for mental health prediction via on- line text data.Proc. ACM Interact. Mob. Wearable Ubiquitous Technol.8, 1–32; 10.1145/3643540 (2024)

View on Pith doi:10.1145/3643540 2024
[5]

Lamichhane, B. Evaluation of ChatGPT for NLP-based mental health applications. Preprint at https://doi.org/10.48550/arXiv.2303.15727(2023)

View on Pith Pith review arXiv doi:10.48550/arxiv.2303.15727 2023
[6]

Ganesan, A.V. et al. Explaining GPTs’ schema of depression: a machine behavior analysis.J. Psychopathol. Clin. Sci.; 10.1037/abn0001144 (2026)

View on Pith doi:10.1037/abn0001144 2026
[7]

Hollinsworth, O.J., Tigges, C., Geiger, A. & Nanda, N. Language models linearly represent sentiment. InProc. 7th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP58–87; 10.18653/v1/2024.blackboxnlp-1.5 (Association for Computational Linguistics, 2024)

View on Pith doi:10.18653/v1/2024.blackboxnlp-1.5 2024
[8]

Di Palma, D. et al. LLaMAs have feelings too: unveiling sentiment and emotion representations in LLaMA models through probing. InProc. 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)6124–6142; 10.18653/v1/2025.acl-long.306 (Association for Computational Linguistics, 2025)

View on Pith doi:10.18653/v1/2025.acl-long.306
[9]

Tak, A.N. et al. Mechanistic interpretability of emotion inference in large language mod- els. InFindings of the Association for Computational Linguistics: ACL 202513090–13120; 10.18653/v1/2025.findings-acl.679 (Association for Computational Linguistics, 2025). 18

View on Pith doi:10.18653/v1/2025.findings-acl.679 2025
[10]

American Psychiatric Association.Diagnostic and Statistical Manual of Mental Disorders5th edn; 10.1176/appi.books.9780890425596 (American Psychiatric Publishing, 2013)

View on Pith doi:10.1176/appi.books.9780890425596 2013
[11]

World Health Organization.The ICD-10 Classification of Mental and Behavioural Disorders (World Health Organization, 1992)

View on Pith 1992
[12]

Hamilton, M. A rating scale for depression.J. Neurol. Neurosurg. Psychiatry23, 56–62; 10.1136/jnnp.23.1.56 (1960)

View on Pith doi:10.1136/jnnp.23.1.56 1960
[13]

Montgomery, S.A. & Åsberg, M. A new depression scale designed to be sensitive to change.Br. J. Psychiatry134, 382–389; 10.1192/bjp.134.4.382 (1979)

View on Pith doi:10.1192/bjp.134.4.382 1979
[14]

Kroenke, K., Spitzer, R.L. & Williams, J.B.W. The PHQ-9: validity of a brief depression severity measure.J. Gen. Intern. Med.16, 606–613; 10.1046/j.1525-1497.2001.016009606.x (2001)

View on Pith arXiv 2001
[15]

Styron, W.Darkness Visible: A Memoir of Madness(Random House, 1990)

View on Pith 1990
[16]

Gotlib, I.H. & Hammen, C.L. (eds)Handbook of Depression3rd edn (Guilford Press, 2014)

View on Pith 2014
[17]

Bao, E., Pérez, A. & Parapar, J. ReDSM5: a Reddit dataset for DSM-5 depression detection. In Proc. 34th ACM International Conference on Information and Knowledge Management (CIKM ’25)6323–6327; 10.1145/3746252.3761610 (ACM, 2025)

View on Pith arXiv
[18]

Shafer, A.B. Meta-analysis of the factor structures of four depression questionnaires: Beck, CES-D, Hamilton, and Zung.J. Clin. Psychol.62, 123–146; 10.1002/jclp.20213 (2006)

View on Pith doi:10.1002/jclp.20213 2006
[19]

Asai, A.etal.HappyDB:acorpusof100,000crowdsourcedhappymoments.InProc. 11th Interna- tional Conference on Language Resources and Evaluation (LREC 2018); 10.63317/2vs32uubt6id (European Language Resources Association, 2018)

View on Pith doi:10.63317/2vs32uubt6id 2018
[20]

Gemma Team. Gemma 3 technical report. Preprint athttps://doi.org/10.48550/arXiv.2 503.19786(2025)

View on Pith doi:10.48550/arxiv.2 2025
[21]

Wolf, T. et al. Transformers: state-of-the-art natural language processing. InProc. 2020 Conference on Empirical Methods in Natural Language Processing: System Demonstrations 38–45; 10.18653/v1/2020.emnlp-demos.6 (Association for Computational Linguistics, 2020)

View on Pith doi:10.18653/v1/2020.emnlp-demos.6 2020
[22]

Paszke, A. et al. PyTorch: an imperative style, high-performance deep learning library. In Advances in Neural Information Processing Systems 328024–8035; 10.48550/arXiv.1912.01703 (Curran Associates, 2019)

View on Pith Pith review arXiv doi:10.48550/arxiv.1912.01703 1912
[23]

Xiao, G., Tian, Y., Chen, B., Han, S. & Lewis, M. Efficient streaming language models with attention sinks. InInternational Conference on Learning Representations (ICLR 2024); 10.48550/arXiv.2309.17453 (2024)

View on Pith Pith review arXiv doi:10.48550/arxiv.2309.17453 2024
[24]

Sun, M., Chen, X., Kolter, J.Z. & Liu, Z. Massive activations in large language models. In Conference on Language Modeling (COLM 2024); 10.48550/arXiv.2402.17762 (2024)

View on Pith Pith review arXiv doi:10.48550/arxiv.2402.17762 2024
[25]

Ansel, J. et al. PyTorch 2: faster machine learning through dynamic Python bytecode transfor- mation and graph compilation. InProc. 29th ACM International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS ’24)Vol. 2, 929–947; 10.1145/3620665.3640366 (ACM, 2024). 19

View on Pith arXiv
[26]

Stone, C. & Zhu, F. ConnorStoneAstro/pted: v1.4.1.Zenodo; 10.5281/zenodo.19903244 (2026)

View on Pith doi:10.5281/zenodo.19903244 2026
[27]

Székely, G.J. & Rizzo, M.L. Energy statistics: a class of statistics based on distances.J. Stat. Plan. Inference143, 1249–1272; 10.1016/j.jspi.2013.03.018 (2013)

View on Pith doi:10.1016/j.jspi.2013.03.018 2013
[28]

Virtanen, P. et al. SciPy 1.0: fundamental algorithms for scientific computing in Python.Nat. Methods17, 261–272; 10.1038/s41592-019-0686-2 (2020)

View on Pith doi:10.1038/s41592-019-0686-2 2020
[29]

Anderson, M.J. A new method for non-parametric multivariate analysis of variance.Austral Ecol.26, 32–46; 10.1046/j.1442-9993.2001.01070.x (2001)

View on Pith arXiv 2001
[30]

Anderson, M.J. Distance-based tests for homogeneity of multivariate dispersions.Biometrics 62, 245–253; 10.1111/j.1541-0420.2005.00440.x (2006)

View on Pith arXiv 2006
[31]

Aton, M. et al. scikit-bio: a fundamental Python library for biological omic data analysis.Nat. Methods23, 274–276; 10.1038/s41592-025-02981-z (2026)

View on Pith doi:10.1038/s41592-025-02981-z 2026
[32]

Grand, G., Blank, I.A., Pereira, F. & Fedorenko, E. Semantic projection recovers rich human knowledge of multiple object features from word embeddings.Nat. Hum. Behav.6, 975–987; 10.1038/s41562-022-01316-8 (2022)

View on Pith doi:10.1038/s41562-022-01316-8 2022
[33]

Park, K., Choe, Y.J., Jiang, Y. & Veitch, V. The geometry of categorical and hierarchical concepts in large language models. InInternational Conference on Learning Representations (ICLR 2025); 10.48550/arXiv.2406.01506 (2025)

View on Pith Pith review arXiv doi:10.48550/arxiv.2406.01506 2025
[34]

Weber, S. et al. Using a fine-tuned large language model for symptom-based depression evaluation. npj Digit. Med.8, 598; 10.1038/s41746-025-01982-8 (2025)

View on Pith doi:10.1038/s41746-025-01982-8 2025
[35]

Panickssery, N. et al. Steering Llama 2 via contrastive activation addition. InProc. 62nd Annual Meeting of the Association for Computational Linguistics15504–15522; 10.18653/v1/2024.acl- long.828 (Association for Computational Linguistics, 2024)

View on Pith doi:10.18653/v1/2024.acl- 2024
[36]

Zou, A. et al. Representation engineering: a top-down approach to AI transparency. Preprint at https://doi.org/10.48550/arXiv.2310.01405(2023)

View on Pith Pith review arXiv doi:10.48550/arxiv.2310.01405 2023
[37]

Parmigiani, S. et al. Reliability and validity of transcranial magnetic stimulation–elec- troencephalography biomarkers.Biol. Psychiatry Cogn. Neurosci. Neuroimaging8, 805–814; 10.1016/j.bpsc.2022.12.005 (2023)

View on Pith doi:10.1016/j.bpsc.2022.12.005 2023
[38]

Huang, Y. et al. Naturalistic acute pain states decoded from neural and facial dynamics.Nat. Commun.16, 4371; 10.1038/s41467-025-59756-5 (2025)

View on Pith doi:10.1038/s41467-025-59756-5 2025
[39]

Kakusa, B. et al. Distinct neural temporal architectures encode rapid social expressions and sustained internal mood states. Preprint atbioRxiv; 10.64898/2025.12.20.692681 (2025)

View on Pith doi:10.64898/2025.12.20.692681 2025
[40]

Cline, C.C. & Keller, C.J. Personalizing brain stimulation for psychiatric disorders: from circuits to closed-loop control.Am. J. Psychiatry183, 169–179; 10.1176/appi.ajp.20251327 (2026). 20 Supplementary Material Core Clinical Symptom Descriptions Natural-language descriptions of depression symptoms, organized by clinical assessment instru- ment and grou...

View on Pith doi:10.1176/appi.ajp.20251327 2026

Discussion

Disagree with the review, or know something it missed? Leave a comment, highlight any passage above to comment on it, or challenge this review formally.

[1] [1]

Fried, E.I. & Nesse, R.M. Depression is not a consistent syndrome: an investigation of unique symptom patterns in the STAR*D study.J. Affect. Disord.172, 96–102; 10.1016/j.jad.2014.10.010 (2015)

View on Pith doi:10.1016/j.jad.2014.10.010 2015

[2] [2]

Chevance, A. et al. Identifying outcomes for depression that matter to patients, informal caregivers, and health-care professionals: qualitative content analysis of a large international online survey.Lancet Psychiatry7, 692–702; 10.1016/S2215-0366(20)30191-7 (2020)

View on Pith doi:10.1016/s2215-0366(20)30191-7 2020

[3] [3]

Yang, K. et al. MentaLLaMA: interpretable mental health analysis on social media with large language models. InProc. ACM Web Conference 2024 (WWW ’24)4489–4500; 10.1145/3589334.3648137 (ACM, 2024)

View on Pith arXiv 2024

[4] [4]

Xu, X. et al. Mental-LLM: leveraging large language models for mental health prediction via on- line text data.Proc. ACM Interact. Mob. Wearable Ubiquitous Technol.8, 1–32; 10.1145/3643540 (2024)

View on Pith doi:10.1145/3643540 2024

[5] [5]

Lamichhane, B. Evaluation of ChatGPT for NLP-based mental health applications. Preprint at https://doi.org/10.48550/arXiv.2303.15727(2023)

View on Pith Pith review arXiv doi:10.48550/arxiv.2303.15727 2023

[6] [6]

Ganesan, A.V. et al. Explaining GPTs’ schema of depression: a machine behavior analysis.J. Psychopathol. Clin. Sci.; 10.1037/abn0001144 (2026)

View on Pith doi:10.1037/abn0001144 2026

[7] [7]

Hollinsworth, O.J., Tigges, C., Geiger, A. & Nanda, N. Language models linearly represent sentiment. InProc. 7th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP58–87; 10.18653/v1/2024.blackboxnlp-1.5 (Association for Computational Linguistics, 2024)

View on Pith doi:10.18653/v1/2024.blackboxnlp-1.5 2024

[8] [8]

Di Palma, D. et al. LLaMAs have feelings too: unveiling sentiment and emotion representations in LLaMA models through probing. InProc. 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)6124–6142; 10.18653/v1/2025.acl-long.306 (Association for Computational Linguistics, 2025)

View on Pith doi:10.18653/v1/2025.acl-long.306

[9] [9]

Tak, A.N. et al. Mechanistic interpretability of emotion inference in large language mod- els. InFindings of the Association for Computational Linguistics: ACL 202513090–13120; 10.18653/v1/2025.findings-acl.679 (Association for Computational Linguistics, 2025). 18

View on Pith doi:10.18653/v1/2025.findings-acl.679 2025

[10] [10]

American Psychiatric Association.Diagnostic and Statistical Manual of Mental Disorders5th edn; 10.1176/appi.books.9780890425596 (American Psychiatric Publishing, 2013)

View on Pith doi:10.1176/appi.books.9780890425596 2013

[11] [11]

World Health Organization.The ICD-10 Classification of Mental and Behavioural Disorders (World Health Organization, 1992)

View on Pith 1992

[12] [12]

Hamilton, M. A rating scale for depression.J. Neurol. Neurosurg. Psychiatry23, 56–62; 10.1136/jnnp.23.1.56 (1960)

View on Pith doi:10.1136/jnnp.23.1.56 1960

[13] [13]

Montgomery, S.A. & Åsberg, M. A new depression scale designed to be sensitive to change.Br. J. Psychiatry134, 382–389; 10.1192/bjp.134.4.382 (1979)

View on Pith doi:10.1192/bjp.134.4.382 1979

[14] [14]

Kroenke, K., Spitzer, R.L. & Williams, J.B.W. The PHQ-9: validity of a brief depression severity measure.J. Gen. Intern. Med.16, 606–613; 10.1046/j.1525-1497.2001.016009606.x (2001)

View on Pith arXiv 2001

[15] [15]

Styron, W.Darkness Visible: A Memoir of Madness(Random House, 1990)

View on Pith 1990

[16] [16]

Gotlib, I.H. & Hammen, C.L. (eds)Handbook of Depression3rd edn (Guilford Press, 2014)

View on Pith 2014

[17] [17]

Bao, E., Pérez, A. & Parapar, J. ReDSM5: a Reddit dataset for DSM-5 depression detection. In Proc. 34th ACM International Conference on Information and Knowledge Management (CIKM ’25)6323–6327; 10.1145/3746252.3761610 (ACM, 2025)

View on Pith arXiv

[18] [18]

Shafer, A.B. Meta-analysis of the factor structures of four depression questionnaires: Beck, CES-D, Hamilton, and Zung.J. Clin. Psychol.62, 123–146; 10.1002/jclp.20213 (2006)

View on Pith doi:10.1002/jclp.20213 2006

[19] [19]

Asai, A.etal.HappyDB:acorpusof100,000crowdsourcedhappymoments.InProc. 11th Interna- tional Conference on Language Resources and Evaluation (LREC 2018); 10.63317/2vs32uubt6id (European Language Resources Association, 2018)

View on Pith doi:10.63317/2vs32uubt6id 2018

[20] [20]

Gemma Team. Gemma 3 technical report. Preprint athttps://doi.org/10.48550/arXiv.2 503.19786(2025)

View on Pith doi:10.48550/arxiv.2 2025

[21] [21]

Wolf, T. et al. Transformers: state-of-the-art natural language processing. InProc. 2020 Conference on Empirical Methods in Natural Language Processing: System Demonstrations 38–45; 10.18653/v1/2020.emnlp-demos.6 (Association for Computational Linguistics, 2020)

View on Pith doi:10.18653/v1/2020.emnlp-demos.6 2020

[22] [22]

Paszke, A. et al. PyTorch: an imperative style, high-performance deep learning library. In Advances in Neural Information Processing Systems 328024–8035; 10.48550/arXiv.1912.01703 (Curran Associates, 2019)

View on Pith Pith review arXiv doi:10.48550/arxiv.1912.01703 1912

[23] [23]

Xiao, G., Tian, Y., Chen, B., Han, S. & Lewis, M. Efficient streaming language models with attention sinks. InInternational Conference on Learning Representations (ICLR 2024); 10.48550/arXiv.2309.17453 (2024)

View on Pith Pith review arXiv doi:10.48550/arxiv.2309.17453 2024

[24] [24]

Sun, M., Chen, X., Kolter, J.Z. & Liu, Z. Massive activations in large language models. In Conference on Language Modeling (COLM 2024); 10.48550/arXiv.2402.17762 (2024)

View on Pith Pith review arXiv doi:10.48550/arxiv.2402.17762 2024

[25] [25]

Ansel, J. et al. PyTorch 2: faster machine learning through dynamic Python bytecode transfor- mation and graph compilation. InProc. 29th ACM International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS ’24)Vol. 2, 929–947; 10.1145/3620665.3640366 (ACM, 2024). 19

View on Pith arXiv

[26] [26]

Stone, C. & Zhu, F. ConnorStoneAstro/pted: v1.4.1.Zenodo; 10.5281/zenodo.19903244 (2026)

View on Pith doi:10.5281/zenodo.19903244 2026

[27] [27]

Székely, G.J. & Rizzo, M.L. Energy statistics: a class of statistics based on distances.J. Stat. Plan. Inference143, 1249–1272; 10.1016/j.jspi.2013.03.018 (2013)

View on Pith doi:10.1016/j.jspi.2013.03.018 2013

[28] [28]

Virtanen, P. et al. SciPy 1.0: fundamental algorithms for scientific computing in Python.Nat. Methods17, 261–272; 10.1038/s41592-019-0686-2 (2020)

View on Pith doi:10.1038/s41592-019-0686-2 2020

[29] [29]

Anderson, M.J. A new method for non-parametric multivariate analysis of variance.Austral Ecol.26, 32–46; 10.1046/j.1442-9993.2001.01070.x (2001)

View on Pith arXiv 2001

[30] [30]

Anderson, M.J. Distance-based tests for homogeneity of multivariate dispersions.Biometrics 62, 245–253; 10.1111/j.1541-0420.2005.00440.x (2006)

View on Pith arXiv 2006

[31] [31]

Aton, M. et al. scikit-bio: a fundamental Python library for biological omic data analysis.Nat. Methods23, 274–276; 10.1038/s41592-025-02981-z (2026)

View on Pith doi:10.1038/s41592-025-02981-z 2026

[32] [32]

Grand, G., Blank, I.A., Pereira, F. & Fedorenko, E. Semantic projection recovers rich human knowledge of multiple object features from word embeddings.Nat. Hum. Behav.6, 975–987; 10.1038/s41562-022-01316-8 (2022)

View on Pith doi:10.1038/s41562-022-01316-8 2022

[33] [33]

Park, K., Choe, Y.J., Jiang, Y. & Veitch, V. The geometry of categorical and hierarchical concepts in large language models. InInternational Conference on Learning Representations (ICLR 2025); 10.48550/arXiv.2406.01506 (2025)

View on Pith Pith review arXiv doi:10.48550/arxiv.2406.01506 2025

[34] [34]

Weber, S. et al. Using a fine-tuned large language model for symptom-based depression evaluation. npj Digit. Med.8, 598; 10.1038/s41746-025-01982-8 (2025)

View on Pith doi:10.1038/s41746-025-01982-8 2025

[35] [35]

Panickssery, N. et al. Steering Llama 2 via contrastive activation addition. InProc. 62nd Annual Meeting of the Association for Computational Linguistics15504–15522; 10.18653/v1/2024.acl- long.828 (Association for Computational Linguistics, 2024)

View on Pith doi:10.18653/v1/2024.acl- 2024

[36] [36]

Zou, A. et al. Representation engineering: a top-down approach to AI transparency. Preprint at https://doi.org/10.48550/arXiv.2310.01405(2023)

View on Pith Pith review arXiv doi:10.48550/arxiv.2310.01405 2023

[37] [37]

Parmigiani, S. et al. Reliability and validity of transcranial magnetic stimulation–elec- troencephalography biomarkers.Biol. Psychiatry Cogn. Neurosci. Neuroimaging8, 805–814; 10.1016/j.bpsc.2022.12.005 (2023)

View on Pith doi:10.1016/j.bpsc.2022.12.005 2023

[38] [38]

Huang, Y. et al. Naturalistic acute pain states decoded from neural and facial dynamics.Nat. Commun.16, 4371; 10.1038/s41467-025-59756-5 (2025)

View on Pith doi:10.1038/s41467-025-59756-5 2025

[39] [39]

Kakusa, B. et al. Distinct neural temporal architectures encode rapid social expressions and sustained internal mood states. Preprint atbioRxiv; 10.64898/2025.12.20.692681 (2025)

View on Pith doi:10.64898/2025.12.20.692681 2025

[40] [40]

Cline, C.C. & Keller, C.J. Personalizing brain stimulation for psychiatric disorders: from circuits to closed-loop control.Am. J. Psychiatry183, 169–179; 10.1176/appi.ajp.20251327 (2026). 20 Supplementary Material Core Clinical Symptom Descriptions Natural-language descriptions of depression symptoms, organized by clinical assessment instru- ment and grou...

View on Pith doi:10.1176/appi.ajp.20251327 2026