ConsistentRFT: Reducing Visual Hallucinations in Flow-based Reinforcement Fine-Tuning

TL;DR

This paper introduces ConsistentRFT, a framework that reduces visual hallucinations in flow-based reinforcement fine-tuning by balancing exploration and maintaining policy consistency, leading to improved model reliability.

Contribution

It proposes a novel framework with DGR and CPGO to address hallucinations in RFT, enhancing global semantics and policy stability.

Findings

49% reduction in low-level hallucinations

38% reduction in high-level hallucinations

5.1% improvement on out-of-domain metrics

Abstract

Reinforcement Fine-Tuning (RFT) on flow-based models is crucial for preference alignment. However, they often introduce visual hallucinations like over-optimized details and semantic misalignment. This work preliminarily explores why visual hallucinations arise and how to reduce them. We first investigate RFT methods from a unified perspective, and reveal the core problems stemming from two aspects, exploration and exploitation: (1) limited exploration during stochastic differential equation (SDE) rollouts, leading to an over-emphasis on local details at the expense of global semantics, and (2) trajectory imitation process inherent in policy gradient methods, distorting the model's foundational vector field and its cross-step consistency. Building on this, we propose ConsistentRFT, a general framework to mitigate these hallucinations. Specifically, we design a Dynamic Granularity Rollout (DGR) mechanism to balance exploration between global semantics and local details by dynamically scheduling different noise sources. We then introduce a Consistent Policy Gradient Optimization (CPGO) that preserves the model's consistency by aligning the current policy with a more stable prior. Extensive experiments demonstrate that ConsistentRFT significantly mitigates visual hallucinations, achieving average reductions of 49\% for low-level and 38\% for high-level perceptual hallucinations. Furthermore, ConsistentRFT outperforms other RFT methods on out-of-domain metrics, showing an improvement of 5.1\% (v.s. the baseline's decrease of -0.4\%) over FLUX1.dev. This is \href{https://xiaofeng-tan.github.io/projects/ConsistentRFT}{Project Page}.