Pretrained and fine-tuned Qwen3 embeddings exhibit measurable alignment with an expert symptom matrix via RSA on Reddit mental-health data, strengthened by fine-tuning at fine-grained levels and larger scale, with residual alignment after VAD/LIWC/topic controls.
Frontiers in Neuroscience , volume =
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Do LLM Embedding Spaces Recover Expert Structure?
Pretrained and fine-tuned Qwen3 embeddings exhibit measurable alignment with an expert symptom matrix via RSA on Reddit mental-health data, strengthened by fine-tuning at fine-grained levels and larger scale, with residual alignment after VAD/LIWC/topic controls.