# WPG-MoE: Weak-Prior-Guided Dense Mixture-of-Experts for User-Level Social Media Depression Detection

arXiv: https://arxiv.org/abs/2607.04350 · HTML: https://mlsanity.com/paper/2607.04350/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Multimodal MoE models, one of 12 directions in Mixture-of-experts.
- Builds on 25 of its 67 references that are in the map.
- First posted on arXiv 5 Jul 2026.

## Placed in

- Mixture-of-experts / [Multimodal MoE models](https://mlsanity.com/moe/multimodal-moe-models/)

## Builds on (most cited first, 25 of 25)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer](https://mlsanity.com/paper/1701.06538/) (2017)
- [Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity](https://mlsanity.com/paper/2101.03961/) (2021)
- [DeBERTaV3: Improving DeBERTa using ELECTRA-Style Pre-Training with Gradient-Disentangled Embedding Sharing](https://mlsanity.com/paper/2111.09543/) (2021)
- [Unifying distillation and privileged information](https://mlsanity.com/paper/1511.03643/) (2015)
- [Depression and Self-Harm Risk Assessment in Online Forums](https://mlsanity.com/paper/1709.01848/) (2017)
- [ChiMed-GPT: A Chinese Medical Large Language Model with Full Training Regime and Better Alignment to Human Preferences](https://mlsanity.com/paper/2311.06025/) (2023)
- [Utilizing Neural Networks and Linguistic Metadata for Early Detection of Depression Indications in Text Sequences](https://mlsanity.com/paper/1804.07000/) (2018)
- [Improving the Generalizability of Depression Detection by Leveraging Clinical Questionnaires](https://mlsanity.com/paper/2204.10432/) (2022)
- [Topic Memory Networks for Short Text Classification](https://mlsanity.com/paper/1809.03664/) (2018)
- [Identifying Depressive Symptoms from Tweets: Figurative Language Enabled Multitask Learning Framework](https://mlsanity.com/paper/2011.06149/) (2020)
- [Named Entity Recognition for Social Media Texts with Semantic Augmentation](https://mlsanity.com/paper/2010.15458/) (2020)
- [medIKAL: Integrating Knowledge Graphs as Assistants of LLMs for Enhanced Clinical Diagnosis on EMRs](https://mlsanity.com/paper/2406.14326/) (2024)
- [Learning Using Privileged Information: SVM+ and Weighted SVM](https://mlsanity.com/paper/1306.3161/) (2013)
- [Psychiatric Scale Guided Risky Post Screening for Early Detection of Depression](https://mlsanity.com/paper/2205.09497/) (2022)
- [Are LLMs effective psychological assessors? Leveraging adaptive RAG for interpretable mental health screening through psychometric practice](https://mlsanity.com/paper/2501.00982/) (2025)
- [DepressionNet: A Novel Summarization Boosted Deep Framework for Depression Detection on Social Media](https://mlsanity.com/paper/2105.10878/) (2021)
- [RSDD-Time: Temporal Annotation of Self-Reported Mental Health Diagnoses](https://mlsanity.com/paper/1806.07916/) (2018)
- [Interpretable Depression Detection from Social Media Text Using LLM-Derived Embeddings](https://mlsanity.com/paper/2506.06616/) (2025)
- [Datasets for Depression Modeling in Social Media: An Overview](https://mlsanity.com/paper/2503.21513/) (2025)
- [Then and Now: Quantifying the Longitudinal Validity of Self-Disclosed Depression Diagnoses](https://mlsanity.com/paper/2206.11155/) (2022)
- [A Gold Standard Dataset and Evaluation Framework for Depression Detection and Explanation in Social Media using LLMs](https://mlsanity.com/paper/2507.19899/) (2025)
- [Hashtag-Guided Low-Resource Tweet Classification](https://mlsanity.com/paper/2302.10143/) (2023)
- [ReDepress: A Cognitive Framework for Detecting Depression Relapse from Social Media](https://mlsanity.com/paper/2509.17991/) (2025)
- [CNSocialDepress: A Chinese Social Media Dataset for Depression Risk Detection and Structured Analysis](https://mlsanity.com/paper/2510.11233/) (2025)

## Abstract (arXiv)

Online social media posts provide scalable signals for early depression screening, and recent studies mainly improve pre-classification evidence through risk-post selection, symptom grounding, and clinically informed feature construction. However, these screening-stage designs often leave final decisions to a single detector, overlooking how users heterogeneously express depressive risk after screening. A monolithic classifier must average across heterogeneous users, which may dilute localized evidence and cause misclassification, especially for non-self-disclosing users. To address this issue, we propose WPG-MoE, a weak-prior-guided dense mixture-of-experts framework built on a shared large language model (LLM) backbone. WPG-MoE derives user-level weak semantic priors to softly route users to experts matched to different evidence layouts. We formulate this process as learning using privileged information (LUPI): rich LLM-extracted structured evidence guides training-time routing, while inference retains only Patient Health Questionnaire-9 (PHQ-9) template screening and the deployable backbone. Experiments on Chinese and English datasets show that WPG-MoE outperforms strong baselines with interpretable routing behavior.