PsychAdapter: Adapting LLM Transformers to Reflect Traits, Personality and Mental Health

2024-12-22

AI Summary

Current large language models (LLMs) generate generic, psychologically ungrounded text due to the absence of explicit representations of personality, age, and psychological states, resulting in low interpretability and poor psychological validity. To address this, we propose PsychAdapter—a lightweight, multi-layer embedding adapter module for Transformer-based LLMs—that injects empirically grounded psychology-to-language mapping rules directly into the model’s底层 architecture. Unlike prompt-based approaches, PsychAdapter enables controllable, interpretable, and context-agnostic generation of psychologically conditioned text without prompt engineering. Built upon LoRA-style parameter-efficient fine-tuning, it supports cross-model adaptation (e.g., GPT-2, Gemma, Llama 3) and explicit modeling of trait–language associations. Experiments demonstrate strong performance: 87.3% accuracy in Big Five personality classification and 96.7% accuracy in depression and life satisfaction detection. Generated outputs exhibit significantly enhanced psychological validity, with zero overhead on LLM context window usage.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Cognitive Modeling & Cognitive Systems: Adaptive Behavior
• Natural Language Processing: (Large) Language Models

Application Category

• User Modeling, Personalization and Recommendation: Psychology-informed user models and recommender systems
• Search and Retrieval-Augmented AI: Large language models for search
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Artificial intelligence-based language generators are now a part of most people's lives. However, by default, they tend to generate"average"language without reflecting the ways in which people differ. Here, we propose a lightweight modification to the standard language model transformer architecture -"PsychAdapter"- that uses empirically derived trait-language patterns to generate natural language for specified personality, demographic, and mental health characteristics (with or without prompting). We applied PsychAdapters to modify OpenAI's GPT-2, Google's Gemma, and Meta's Llama 3 and found generated text to reflect the desired traits. For example, expert raters evaluated PsychAdapter's generated text output and found it matched intended trait levels with 87.3% average accuracy for Big Five personalities, and 96.7% for depression and life satisfaction. PsychAdapter is a novel method to introduce psychological behavior patterns into language models at the foundation level, independent of prompting, by influencing every transformer layer. This approach can create chatbots with specific personality profiles, clinical training tools that mirror language associated with psychological conditionals, and machine translations that match an authors reading or education level without taking up LLM context windows. PsychAdapter also allows for the exploration psychological constructs through natural language expression, extending the natural language processing toolkit to study human psychology.

Problem

• Personalization
• Language Generation
• User Characteristics

Innovation

• PsychAdapter
• Personalized Text Generation
• Machine Learning in Psychology