G-Frame framework trains OmniChem 7B with 79% fewer hallucinations

Researchers developed G-Frame, an adaptive multi-agent framework that integrates Bayesian and team game principles to mitigate hallucinations in lightweight large language models. This approach establishes an automated closed-loop for high-quality data synthesis and model training by forcing the internalization of domain constraints through structured reasoning.

• The framework synthesized a specialized corpus containing 363,045 chains-of-thought and 199,589 question-answer pairs.
• The resulting 7B model, OmniChem, achieves performance parity with GPT 4o mini on custom benchmarks and ChemBench.
• OmniChem exhibits a 79.46% reduction in hallucinations relative to its base architecture.
• The system demonstrates advanced capabilities in molecular design and synthesis planning.

This work establishes a scalable paradigm utilizing adaptive multi-agents to overcome inherent reasoning deficiencies, offering a feasible pathway for accelerating knowledge discovery in specialized scientific fields.