Direct Preference Density Alignment for Conversational Audio Equalization
This work proposes Direct Preference Density Alignment: An alternative framework that removes the need for a learned proxy reward model while strictly preserving the benefits of online reinforcement learning, and enables the combination of the online structural grounding of Group Relative Policy Optimization with the t...