AdaC-GRPO: Towards Mastering Long-Horizon Visual Sudoku Puzzles via Adaptive Curriculum Learning

To this end, we introduce FG-Sudoku, a fine-grained benchmark categorized by grid size and required deductive steps. It comes from a paper posted to arXiv on 12 October 2026. While reinforcement learning with verifiable reward (RLVR) enhances reasoning in large language models, its efficacy in long-horizon multimodal tasks remains underexplored. Visual Sudoku puzzles pose a representative challenge, requiring multi-step deductive reasoning over both local constraints and global dependency chains. Evaluations reveal that frontier MLLMs fail on complex long-range dependencies (e.g., 9*9 puzzles). During post-training, this near-zero success rate causes severe reward sparsity, halting standard RLVR. We propose Adaptive Curriculum Group Relative Policy Optimization (AdaC-GRPO) to overcome this. Unlike rigid or prior-based schedules, AdaC-GRPO dynamically constructs training batches by tracking the model's "Zone of Proximal Development" via online rollout accuracy. Experiments demonstrate that AdaC-GRPO enables Qwen2.5-VL-7B to approach the performance of Gemini-2.5-Flash while bridging 62.2% of the overall accuracy gap to Gemini-2.5-Pro, with its generalizability further validated across InternVL-3.5 and Qwen3-VL.

Why it counts

To this end, we introduce FG-Sudoku, a fine-grained benchmark categorized by grid size and required deductive steps.

Sources

Primary source: primary source

What is not known

This brief does not claim independent replication. Claims that appear only on X and not in the primary source stay unknown.