DRIFT: Transferring Reasoning Priors for Efficient MLLM Fine-Tuning

Multimodal large language models (MLLMs) have made rapid progress, yet their reasoning ability often lags behind strong text-only LLMs.Bridging this gap typically requires large-scale multimodal reasoning data or reinforcement learning, incurring substantial cost.An appealing alternative is parameter-space model merging between reasoning-enhanced LLMs and MLLMs, but we show that naive merging is fragile: its effectiveness varies widely across model families and can significantly degrade performance (e.g., for Qwen-based MLLMs).We propose Directional Reasoning Injection for Fine-Tuning (DRIFT), a lightweight method that transfers reasoning knowledge in the gradient space while preserving multimodal alignment.DRIFT precomputes a reasoning prior from the parameter differences between text-only reasoning experts and multimodal models, and uses it to bias gradients during supervised fine-tuning.This design retains the simplicity of standard SFT pipelines while enabling efficient and stable reasoning transfer.Experiments on multimodal reasoning benchmarks, including Math-Vista and MathVerse, show that DRIFT consistently outperforms naive merging and standard SFT, and matches or surpasses trainingintensive methods with substantially lower data and compute.