Focused Technical Areas

Using CI to Enhance GPAIS Performance and Expand GPAIS Application Boundaries
▪ Neuro-symbolic and CI-enhanced LLM architectures
▪ CI-based preprocessing for GPAIS
▪ Foundation models
▪ Explainability and safety in GPAIS
▪ Low-resource adaptation and efficiency
▪ CI for robust multi-modal and multi-agent AI
▪ Evolutionary fine-tuning and prompt optimization
▪ Structural optimization of LLM for different objectives (alignment, XAI, efficiency)
▪ Large scale transformers and distributed computing strategies to build GPAIS

Using GPAIS to Advance Intelligent, Explainable, and Semantic-Aware CI: Leverage GPAIS to select, configure, improve, and generate CI techniques, including but not limited to:
▪ GPAIS-driven intelligent evolutionary algorithms
▪ GPAIS-driven intelligent fuzzy systems
▪ GPAIS for automated CI algorithm configuration and new CI algorithm design. Automated algorithm construction using LLMtranslated domain knowledge.
▪ GPAIS-driven explainable CI techniques
▪ Design of more efficient CI techniques

Application scenarios for CI+GPAIS
▪ Benchmarks and validation frameworks for GPAIS
▪ Software engineering
▪ Creative design
▪ Robotics
▪ Conversational agents
▪ Bioinformatics
▪ Healthcare diagnosis with CI-enhanced LLMs
▪ Neuro-symbolic architectures in autonomous vehicle navigation
▪ …