DiverseAR: Boosting Diversity in Bitwise Autoregressive Image Generation
This addresses a specific challenge in image generation for researchers and practitioners, but it is incremental as it builds on existing bitwise AR models.
The paper tackled the problem of limited sample diversity in autoregressive generative models with bitwise visual tokenizers, proposing DiverseAR to enhance diversity without sacrificing visual quality, achieving substantial improvements in experiments.
In this paper, we investigate the underexplored challenge of sample diversity in autoregressive (AR) generative models with bitwise visual tokenizers. We first analyze the factors that limit diversity in bitwise AR models and identify two key issues: (1) the binary classification nature of bitwise modeling, which restricts the prediction space, and (2) the overly sharp logits distribution, which causes sampling collapse and reduces diversity. Building on these insights, we propose DiverseAR, a principled and effective method that enhances image diversity without sacrificing visual quality. Specifically, we introduce an adaptive logits distribution scaling mechanism that dynamically adjusts the sharpness of the binary output distribution during sampling, resulting in smoother predictions and greater diversity. To mitigate potential fidelity loss caused by distribution smoothing, we further develop an energy-based generation path search algorithm that avoids sampling low-confidence tokens, thereby preserving high visual quality. Extensive experiments demonstrate that DiverseAR substantially improves sample diversity in bitwise autoregressive image generation.