Comparative Evaluation of Reasoning and Inference in LLM-Based and Diffusion-Based Approaches

Abstract

Large generative models are widely used in artificial intelligence (AI) systems for autonomous data processing and human-computer interactions. AI systems are expected to interact with complex environments and conduct logical reasoning to provide precise problem-solving abilities for generalized and domain-specific problems. To achieve optimal results, different procedures have been developed around the inference process of these large generative models to compose a refined reasoning flow. This paper aims to study the inference process of two primary types of inference models: the large language model (LLM) and the diffusion model. The purpose of this study is to understand the primary conditions for achieving advanced reasoning results by utilizing the inference of these two models, and to identify the gaps and limitations. This paper covers three primary topics: Reinforcement Learning for LLMs, test-time conditional inference, and diffusion-based inference models. These three topics help in understanding the training, implementation, and adaptation processes of recent LLM-based AI systems. Meanwhile, the diffusion-based model can provide alternative solutions for popular reasoning frameworks. Finally, this paper reviews and evaluates recent advances, comparing their commonalities and differences. In summary, nine recommendations are made for future research in building reasoning AI systems.