Omni-AutoThink: Adaptive Multimodal Reasoning via Reinforcement Learning
This addresses the issue of inefficient or inadequate reasoning in multimodal AI systems, which is incremental as it builds on existing Omni models with adaptive fine-tuning and reinforcement learning.
The paper tackled the problem of rigid reasoning behaviors in Omni models by proposing Omni-AutoThink, an adaptive reasoning framework that dynamically adjusts reasoning depth based on task difficulty, resulting in significant performance improvements on a comprehensive multimodal benchmark.
Recent advances in Omni models have enabled unified multimodal perception and generation. However, most existing systems still exhibit rigid reasoning behaviors, either overthinking simple problems or failing to reason when necessary. To address this limitation, we propose Omni-AutoThink, a novel adaptive reasoning framework that dynamically adjusts the model's reasoning depth according to task difficulty. Our framework comprises two stages: (1) an Adaptive Supervised Fine-Tuning (Adaptive SFT) stage, which endows the Omni model with fundamental reasoning capability using large-scale reasoning-augmented data, and (2) an Adaptive Reinforcement Learning (Adaptive GRPO) stage, which optimizes reasoning behaviors based on task complexity and reward feedback. We further construct a comprehensive adaptive reasoning benchmark that spans text-only, text-audio, text-visual, and text-audio-visual modalities, providing both training and evaluation splits for multimodal reasoning assessment. Experimental results demonstrate that our proposed framework significantly improves adaptive reasoning performance compared to previous baselines. All benchmark data and code will be publicly released.