We introduce SAIL-RL, a reinforcement learning (RL) post-training framework that enhances the reasoning capabilities of multimodal large language models (MLLMs) by teaching them when and how to think. Existing approaches are limited by outcome-only supervision, which rewards correct answers without ensuring sound reasoning, and by uniform thinking strategies, which often lead to overthinking on simple tasks and underthinking on complex ones. SAIL-RL addresses these challenges with a dual reward system: the Thinking Reward, which evaluates reasoning quality through factual grounding, logical coherence, and answer consistency, and the Judging Reward, which adaptively determines whether deep reasoning or direct answering is appropriate. Experiments on the state-of-the-art SAIL-VL2 show that SAIL-RL improves reasoning and multimodal understanding benchmarks at both 4B and 8B scales, achieving competitive performance against commercial closed-source models such as GPT-4o, and substantially reduces hallucinations, establishing it as a principled framework for building more reliable and adaptive MLLMs. The code will be available at https://github.com/BytedanceDouyinContent/SAIL-RL.

翻译：我们提出了SAIL-RL，一种强化学习（RL）后训练框架，通过教导多模态大语言模型（MLLMs）何时及如何思考来增强其推理能力。现有方法受限于仅以结果为导向的监督（仅奖励正确答案而无法确保推理过程的合理性）以及统一的思考策略（常导致简单任务上过度思考而复杂任务上思考不足）。SAIL-RL通过双奖励系统应对这些挑战：思考奖励，通过事实依据、逻辑连贯性和答案一致性来评估推理质量；以及判断奖励，自适应地决定是进行深度推理还是直接作答更为合适。在先进的SAIL-VL2模型上的实验表明，SAIL-RL在4B和8B规模上均提升了推理和多模态理解基准性能，达到了与GPT-4o等商业闭源模型相竞争的水平，并显著减少了幻觉现象，从而确立为一个构建更可靠、更具适应性MLLMs的原则性框架。代码将在https://github.com/BytedanceDouyinContent/SAIL-RL 发布。