DiagramNet supplies a new multimodal dataset and progressive training pipeline with decoupled multi-agent workflow, allowing a 3B model to outperform GPT-5, Claude-Sonnet-4, and Gemini-2.5-Pro by over 2x on system-level diagram tasks while generalizing to other benchmarks.
Boosting MLLM Reasoning with Text-Debiased Hint-GRPO
5 Pith papers cite this work. Polarity classification is still indexing.
5
Pith papers citing it
abstract
MLLM reasoning has drawn widespread research for its excellent problem-solving capability. Current reasoning methods fall into two types: PRM, which supervises the intermediate reasoning steps, and ORM, which supervises the final results. Recently, DeepSeek-R1 has challenged the traditional view that PRM outperforms ORM, which demonstrates strong generalization performance using an ORM method (i.e., GRPO). However, current MLLM's GRPO algorithms still struggle to handle challenging and complex multimodal reasoning tasks (e.g., mathematical reasoning). In this work, we reveal two problems that impede the performance of GRPO on the MLLM: Low data utilization and Text-bias. Low data utilization refers to that GRPO cannot acquire positive rewards to update the MLLM on difficult samples, and text-bias is a phenomenon that the MLLM bypasses image condition and solely relies on text condition for generation after GRPO training. To tackle these problems, this work proposes Hint-GRPO that improves data utilization by adaptively providing hints for samples of varying difficulty, and text-bias calibration that mitigates text-bias by calibrating the token prediction logits with image condition in test-time. Experiment results on three base MLLMs across eleven datasets demonstrate that our proposed methods advance the reasoning capability of original MLLM by a large margin, exhibiting superior performance to existing MLLM reasoning methods. Our code is available at https://github.com/hqhQAQ/Hint-GRPO.
years
2026 5representative citing papers
AdaPrefix-GRPO treats solution-prefix length as a feedback controller targeting 50% rollout success rate during GRPO training, then anneals to zero prefix, yielding 1.6–2.1× accuracy gains over vanilla GRPO at matched compute on hard math.
MHPR is a multidimensional benchmark for LVLM human-centric perception-reasoning with C-RD, SFT-D, RL-D, T-D data tiers and ACVG pipeline, showing training gains on Qwen2.5-VL-7B to near-parity with larger models.
ECHO jointly optimizes policy and critic via co-evolution, cascaded rollouts, and saturation-aware shaping to deliver non-stale feedback and higher success in open-world LLM agent RL.
A GRPO framework that treats thinking as a tool call and uses dual-level regulation so multimodal models learn when to reason versus answer directly.
citing papers explorer
-
DiagramNet: An End-to-End Recognition Framework and Dataset for Non-Standard System-Level Diagrams
DiagramNet supplies a new multimodal dataset and progressive training pipeline with decoupled multi-agent workflow, allowing a 3B model to outperform GPT-5, Claude-Sonnet-4, and Gemini-2.5-Pro by over 2x on system-level diagram tasks while generalizing to other benchmarks.
-
Max Out GRPO Signal: Adaptive Trace Prefix Control for Hard Reasoning Problems
AdaPrefix-GRPO treats solution-prefix length as a feedback controller targeting 50% rollout success rate during GRPO training, then anneals to zero prefix, yielding 1.6–2.1× accuracy gains over vanilla GRPO at matched compute on hard math.
-
MHPR: Multidimensional Human Perception and Reasoning Benchmark for Large Vision-Languate Models
MHPR is a multidimensional benchmark for LVLM human-centric perception-reasoning with C-RD, SFT-D, RL-D, T-D data tiers and ACVG pipeline, showing training gains on Qwen2.5-VL-7B to near-parity with larger models.
-
No More Stale Feedback: Co-Evolving Critics for Open-World Agent Learning
ECHO jointly optimizes policy and critic via co-evolution, cascaded rollouts, and saturation-aware shaping to deliver non-stale feedback and higher success in open-world LLM agent RL.
-
Switch-Reasoner: Learn When to Think in Multitask Mixtures via Reinforcement Learning
A GRPO framework that treats thinking as a tool call and uses dual-level regulation so multimodal models learn when to reason versus answer directly.