psytechlab uses NLP and LLMs for CLPsych 2026 summarization

psytechlab participated in the CLPsych Shared Task 2026, applying Natural Language Processing methods and Large Language Models to analyze social media text for mental health states and well-being. The team utilized techniques including Long Short-Term Memory (LSTM) and BERT-based models to perform self-state analysis and summarization.

• Achieved one of the top Consistency and Contradiction scores in the summarization task.
• Secured middle-level results for the remaining tasks within the shared task.
• Contributed to improving mental health support systems through the development of these estimation tools.

The work demonstrates the utility of automated analysis tools for processing social media data, with code made available via GitHub.