COGN-QWEN4B-4bit: Cognitive Feature Prediction Model

COGN-QWEN4B-4bit is a specialized 4 billion parameter language model developed by FritzStack, built upon the Qwen architecture. Its primary function is to analyze textual input and predict specific cognitive features, making it a unique tool in the domain of psychological text analysis.

Key Capabilities

• Cognitive Bias Detection: Identifies and classifies attention bias, interpretation bias, and memory bias (e.g., Negative).
• Rumination Analysis: Distinguishes types of rumination, such as 'Brooding'.
• Specialized Fine-tuning: Optimized for mental health-related text analysis, focusing on psychological indicators rather than general language tasks.
• Easy Integration: Provides a straightforward Python interface (CognitivePredictor) for quick deployment and use.

Good For

• Mental Health Research: Analyzing large text datasets for patterns in cognitive biases and rumination.
• Psychological Assessment Tools: Integrating into applications that require automated detection of specific psychological states from user-generated text.
• Content Moderation: Identifying text that may indicate certain cognitive patterns relevant to user well-being.