Integral AI developed Collective Agency, a holistic alignment value, and a self-improving Dynamic Alignment framework for Large Language Models. This approach significantly improves an LLM's adherence to Collective Agency without degrading its general language processing abilities, demonstrating a scalable method for aligning future AI systems.
There are no more papers matching your filters at the moment.