salihfurkaan/Affectra-8B
Affectra-8B by salihfurkaan is an 8 billion parameter dense transformer model instruction-tuned for emotional intelligence. It emphasizes empathetic understanding, tone appropriateness, and supportive conversational behavior. The model achieves state-of-the-art performance for its size on EQ-Bench (v2), making it suitable for emotionally aware conversational agents and social reasoning studies.
Loading preview...
Affectra-8B: Emotionally Intelligent Dialogue Model
Affectra-8B is an 8 billion parameter, instruction-tuned dense transformer model developed by salihfurkaan, specifically designed to enhance emotional intelligence in large language models. While many LLMs excel at reasoning, Affectra-8B prioritizes affective understanding and empathetic language generation, addressing the common limitation of social nuance in AI.
Key Capabilities & Design:
- Empathetic Dialogue: Optimized for emotional validation, supportive responses, and socially appropriate conversational tone.
- Structured Representation: Employs a unique design that integrates linguistic stability, social reasoning, and emotional expressiveness across its layers.
- SLERP Optimization: Utilizes Spherical Linear Interpolation (SLERP) for smooth behavioral transitions and stable generation.
Performance & Use Cases:
Affectra-8B demonstrates strong performance on emotional intelligence tasks, achieving a score of 69.9 on EQ-Bench (v2), outperforming larger models like Qwen2.5-7B-Instruct and Llama-3-8B-Instruct in this domain. This makes it particularly well-suited for:
- Emotionally aware conversational agents and chatbots.
- Supportive dialogue systems (non-clinical applications).
- Human-centered AI research and affective computing studies.
Limitations:
It is crucial to note that Affectra-8B is not a substitute for professional advice (medical, psychological, or legal). Human oversight is recommended for sensitive applications, as emotional fluency does not guarantee factual correctness.