GFlowGR: Fine-tuning Generative Recommendation Frameworks with Generative Flow Networks

ArXi:2506.16114v3 Announce Type: replace-cross Generative recommendations (GR), which usually include item tokenizers and generative Large Language Models (LLMs), have nstrated remarkable success across a wide range of scenarios. The majority of existing research efforts primarily concentrate on developing powerful item tokenizers or advancing LLM decoding strategies to attain superior performance. However, the critical fine-tuning step in GR frameworks, which is essential for adapting LLMs to recommendation data, remains largely unexplored.