Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This study investigates how large language models (LLMs) represent depressive symptoms by analyzing the internal activations of Gemma-3-27B-PT using mechanistic interpretability techniques. The researchers found that symptom groups were distinctly separated in the model's layer 21, and they successfully projected naturalistic text onto constructed Symptom Vectors, which preserved the rank ordering of symptoms as annotated by clinicians. Notably, a single depression vector was identified that effectively differentiates between depressive and non-depressive text, achieving an AUC of 0.789, thus providing a foundation for interpretable tools in depression assessment.
A single depression vector in a large language model can accurately distinguish between depressive and non-depressive speech, paving the way for more interpretable mental health assessments.
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.