RL for VLMs

RL: Reinforcement Learning · VLM: Vision-Language Model

Follow this topic Ask about this topic

Momentum

36 papers in the last four weeks, up 227% on the four weeks before. 0.4% of all new papers.

Jul 13Week of Sep 28

Latest papers 287

Aug 2, 2026cs.CV

Credit the Right Box: Marginal Contribution Assignment for Structured Visual Perception

Multimodal Large Language Models (MLLMs) are increasingly expected to solve structured perception tasks that require visual recognition, language-to-object binding, object cardinality preservation, and precisely localized grounding and segmentation outputs. However, existing group-relative reinforcement learning methods provide only response-level supervision, creating a granularity mismatch for structured multi-object prediction: a single advantage is broadcast to all tokens in a response, without distinguishing individual box contributions. To address this mismatch, we propose MCR-GRPO, a marginal contribution assignment framework that derives box-level credit directly from each sampled response. Specifically, Marginal Contribution Reward (MCR) estimates each predicted box's contribution through a leave-one-out comparison, measuring how the matched set value changes when the box is removed from the response. After within-response normalization, records that improve the set value receive positive credit, while redundant or harmful ones are suppressed. To make marginal attribution stable and informative, we further introduce a Continuous Matched Set Value Evaluator that integrates permutation-invariant matching, count-aware normalization, and graded localization. MCR-GRPO maps normalized box-level marginal advantages to the token spans that generated each box, preserving GRPO's response-level comparison while enabling box-aware optimization of structured multi-object grounding. Experiments across REC, DOD, segmentation, and counting benchmarks show state-of-the-art performance over prior GRPO-based baselines.

Xinheng Han, Jianfei Wang, Yu Chen +4

School of Automation, Beijing Institute of Technology · Amap, Alibaba Group

Multimodal Grounding Group Relative Policy Optimization Credit Assignment Fine-Grained Visual Perception RL for VLMs

Aug 2, 2026cs.RO

RL Bootstrapping of OpenVLA-OFT for a Novel Robot Embodiment

Adapting a pretrained vision-language-action (VLA) policy to a new robot usually assumes embodiment-specific demonstrations. This assumption is especially restrictive for custom robots whose morphology differs strongly from the manipulators seen in large robot datasets. We study a harder setting: zero-demo embodiment alignment of OpenVLA-OFT on a cable-driven parallel robot (CDPR) with a simple gripper and a previously unseen control interface. Instead of supervised fine-tuning, we use reinforcement learning in simulation with dense geometric rewards computed from simulator state. The training is performed in two stages: a PPO stage for directional motion primitives, followed by GRPO continuation from the PPO checkpoint with an expanded instruction space that includes object-conditioned commands. On the four shared directional instructions, the average held-out success rate improves from 34.25% after PPO to 53.50% after PPO

\rightarrow

GRPO, with especially large gains on \texttt{move left} and \texttt{move backward}. In the GRPO stage we additionally introduce \texttt{move to <object>} over eight target objects and obtain 39/400 = 9.75% strict success, while qualitative rollouts frequently show correct target-directed approach behavior before late-stage instability. Compared with prior OpenVLA and OpenVLA-OFT results, which rely on demonstration datasets and mostly standard rigid-arm embodiments, our method uses no embodiment-specific dataset at all. The results do not yet establish robust manipulation, but they provide stronger evidence that RL-only bootstrapping can create the first usable language-conditioned controller for a genuinely novel embodiment.

Damir Nurtdinov, Alexei Kornaev, Alexander Maloletov

Research Center of the Artificial Intelligence Institute Innopolis University Innopolis, Russia

Robot Skill Learning Cross-Embodiment Robot Learning Robotic RL Robotics Language-Conditioned Robot Control RL for VLMs

Aug 1, 2026cs.CV

LUT: Latent Utility Training for Visual Reasoning

Multimodal large language models have advanced visual understanding, yet perception-intensive reasoning remains challenging. Recent latent visual reasoning methods introduce hidden-space computation before answering, but they often rely on costly intermediate supervision, such as bounding boxes, sketches, or interleaved rationales. These strategies focus on how latent states should be shaped, but do not explicitly assess whether the latent is useful for the final answer. We propose LUT, a latent reasoning framework trained with only standard VQA pairs. LUT centers training on Latent Utility at two levels. At the trajectory level, we propose Utility-Aware Latent Distillation SFT, which explores answer-relevant latent trajectories, selects qualified trajectories by their information gain, and distills more reliable and learnable supervision through curriculum learning. At the step level, we propose Latent Attribution Policy Optimization, which uses answer-to-latent attribution to differentially optimize latent steps during reinforcement learning. Experiments on perception-intensive visual reasoning benchmarks show that LUT outperforms previous latent reasoning methods and remains competitive with latent-text interleaved methods with lower annotation cost.

Jiaxuan Kang, Siyu Chen, Mingda Li +6

Tencent PCG

Visual Question Answering Large Vision-Language Models Latent Visual Reasoning VLM Reasoning RL for VLMs

Aug 1, 2026cs.CV

Element-Aware Group Learning for E-Commerce Image Generation

Recent advances in image generation and editing have made prompt quality a key bottleneck for e-commerce creatives. Vision-language models (VLMs) can generate image-editing prompts from product images and metadata, but further improving their prompt-writing capabilities requires post-training with feedback from the generated images. Group Relative Policy Optimization (GRPO) is a natural framework for such outcome-level reward optimization. However, it assigns credit only at the full-prompt level, even though image quality often depends on specific design elements such as composition, background, and the presentation of selling points. Existing fine-grained credit assignment methods typically require step-level supervision or learned critics. To address this, we propose EAGLE-GRPO (Element-Aware Group Learning for E-Commerce Image Generation), which decomposes the group-centered reward over predefined elements. We cast element-level credit assignment as a kernel ridge regression problem and derive a closed-form solution, without additional rollouts or separate credit-assignment models. This yields interpretable per-element advantages and more precise policy updates. Experiments show that EAGLE-GRPO sustains performance gains over more training steps before plateauing and generates prompts that produce higher-quality e-commerce images than competitive VLM prompt-writing baselines.

Jingtong Chen, Jiahui Wang, Xue Zhao +2

Group Relative Policy Optimization Image Generation RL for Image Editing RL for VLMs Fine-Grained Credit Assignment

Jul 31, 2026cs.RO

WCM: A World Critic Model for Vision-Language-Action Reinforcement Learning

Reinforcement learning (RL) post-training of Vision-Language-Action (VLA) models has shown strong promise for robotic manipulation. Among RL methods, critic-based approaches rely on a value estimator that predominantly operates on single-frame observations or single-frame VLM backbone latents, which is a fundamental mismatch with the partially observable nature of robot control. A naive approach to incorporate observation history into the critic incurs exponential complexity with high-dimensional visual space, and still fails because pure scalar-return regression provides insufficient supervision for learning cross-temporal dynamics. We identify the root cause as a state approximation problem: without an explicit world modeling objective, the critic's representation cannot capture the temporal structure needed for accurate value estimation. To address this, we propose the World Critic Model (WCM), built on a lightweight LeJEPA architecture; WCM jointly predicts future latent state and estimates values, such that the critic's representation is explicitly trained to capture temporal dynamics rather than merely regress scalar returns. WCM integrates seamlessly into both on-policy and off-policy training pipelines and is compatible with state-of-the-art VLA backbones including Pi0, Pi0.5, and OpenVLA-OFT. Extensive experiments on 149 tasks across four benchmarks demonstrate that WCM consistently achieves state-of-the-art performance in both in-distribution and out-of-distribution settings, with particularly strong generalization gains. We further validate WCM on seven real-world manipulation tasks using OpenVLA-OFT and Pi0.5 with off-policy RL, confirming stable deployment across diverse settings.

Senyu Fei, Xiaopeng Yu, Siyin Wang +3

1Tongji University · 3Fudan University · 2Shanghai Innovation Institute

World Model Learning Vision-Language-Action Models RL for VLMs

Jul 30, 2026cs.CV

Beacon: Knowing When and How to Perform Agentic Visual Reasoning

The fundamental goal of agentic visual reasoning is to improve the success rate of multimodal large language models (MLLMs) on complex tasks. We rethink agentic visual reasoning through two key dimensions of tool use: Mode Adaptiveness and Tool Effect. Mode Adaptiveness characterizes whether an MLLM recognizes when tools are necessary and invokes them accordingly, avoiding unnecessary computational overhead while improving performance on problems requiring tool assistance. Tool Effect characterizes whether tools extend the model's capabilities on problems unsolvable through tool-free reasoning without introducing errors on problems it can already solve. Our analysis quantifies these properties and reveals that existing models exhibit limited Mode Adaptiveness, while tool-use gains on hard examples are largely offset by harm on easy ones. Motivated by these observations, we propose Beacon, a novel agentic visual reasoning model trained with supervised fine-tuning (SFT) and reinforcement learning (RL). Its RL stage combines Necessity-Aware Adaptive Reward and Hint-Guided Capability Expansion. Necessity-Aware Adaptive Reward encourages tool-free solutions when they succeed while preserving full reward for successful tool use when tool-free rollouts fail. Hint-Guided Capability Expansion uses verified, answer-free expert hints to recover learning signals from all-wrong rollout groups, aiming to extend tool-use capability on the hardest problems. Across 13 benchmarks, Beacon achieves the highest average score among the evaluated open-source models and ranks first on 11 benchmarks. On five diagnostic benchmarks, it improves the average tool-available accuracy over its tool-free accuracy by 1.96 points and achieves the largest tool-gain minus tool-harm score (+3.14 points). These results show Beacon's advanced performance, Mode Adaptiveness, and the net benefit of tool use.

Qixun Wang, Yang Shi, Letian Cheng +11

Peking University · Kling Team · HKUST(GZ) +3

Tool Use in VLMs RL for VLMs

Jul 30, 2026cs.CV

Can Vision-Language Models Reason about AI Edits in Images?

Detection and localization of AI-tampered images are critical for trustworthy AI, yet modern generative models have made such manipulations increasingly difficult to identify. While traditional binary classifiers can detect image tampering, they lack interpretability and generalization. Vision-Language Models (VLMs) offer a promising alternative due to their strong visual understanding and reasoning capabilities; however, existing approaches typically rely on supervised finetuning with curated explanations rather than exploiting their inherent reasoning capabilities. In this work, we investigate whether VLMs can be trained to reason about AI-generated image edits using reinforcement learning (RL) rather than explicit reasoning supervision. Motivated by the success in Group Relative Policy Optimization (GRPO), an RL technique that incentivizes the model to reason by generating thinking traces prior to giving the final answer, we propose a GRPO-based training framework that utilizes simple accuracy and format rewards. Given an input image, the model produces a structured reasoning trace and predicts whether the image has been tampered with. A lightweight segmentation model is then guided by the reasoning output to generate pixel-level localization masks. Experiments across multiple image manipulation datasets demonstrate that our approach achieves competitive detection and localization performance compared to state-of-the-art image forgery detectors, despite requiring substantially weaker supervision. We introduce effective intersection over union (eff-IoU), a unified metric to jointly evaluate detection and localization. These results suggest that reinforcement learning provides an effective and scalable mechanism for teaching VLMs to reason about AI-generated content.

Darsha Udayanga, Pin-Yu Chen, Payel Das +1

1Rensselaer Polytechnic Institute · 2IBM Research

Image Forgery Detection Image Forgery Localization VLM Reasoning RL for VLMs

Jul 30, 2026cs.CV

FaithEyes: Towards Faithful Tool Use via Multi-Agent Process-Image Self-Verification

Agentic vision-language models (VLMs), which interleave textual reasoning with explicit tool calls such as cropping and code-based image manipulation, have emerged as a compelling paradigm for reliable and interpretable multi-modal reasoning. However, recent studies have revealed that such models often use tools unfaithfully. Many process images are irrelevant to the question (e.g., the crops miss the queried target), yet the tool call still receives full credit and the model still answers correctly. Such decorative or misaligned tool calls waste computation and reveal that the model does not faithfully use the evidence it retrieves. This may stem from two limitations of prevailing methods: the tool reward fails to distinguish useful from useless calls, and tool feedback carries no signal of usefulness. To this end, we introduce FaithEyes, a multi-agent self-judging framework. Concretely, we use a VLM to judge whether each process image helps answer the question. The judgement is injected into the reasoning context as part of the tool observation to help subsequent reasoning, and meanwhile is used to scale the tool reward by the helpful-tool ratio to suppress reward hacking. To keep judgement available at evaluation, we further design a multi-agent framework where the model itself serves as a subagent to judge the tool calls from the main agent, eliminating any dependence on external models at inference. Training via a two-stage SFT + RL pipeline on adapted open-source data, FaithEyes attains competitive or superior accuracy across visual perception and reasoning benchmarks, while substantially improving tool faithfulness and reducing inference cost. The homepage is at https://github.com/Mosi-AI/FaithEyes.

Haoqing Wang, Xingrun Xing, Ziheng Li +2

Samsung Research, Beijing, China · Peking University

Tool Use in VLMs VLM Reasoning Faithfulness in VLMs RL for VLMs

Jul 30, 2026cs.CV

One Patch Is Enough: Reinforcement-Optimized Visual Token Grounding for MLLM-Based Scene Text Spotting

Scene text spotting requires high-precision alignment between textual recognition and spatial localization. While visual-token grounding has emerged as a promising formulation for Multimodal Large Language Models (MLLMs), the previous multi-patch paradigm often introduces redundant noise and localization ambiguity, particularly for dense or small text instances. To address this, we propose Single-Patch Text Spotting (SPaTS), a vision-centric framework that routes each text instance through a single anchor visual token and then recovers geometry via full-image refinement. To accurately identify this anchor without oracle labels, we introduce Single-Patch Selective Optimization (SPaSO), a reinforcement learning framework that optimizes discrete visual-token selection using patch-level rewards. To further improve representation robustness and localization precision, we introduce Directional Embedding Alignment (DEA) to suppress unstable norm bias by decoupling feature magnitude and direction, and Patch-Enhanced Decoding (PED) to fuse the routed anchor with language semantics and cross-attend over the full-image feature map for geometry-aware boundary regression beyond coordinate-space surrogates. Extensive experiments demonstrate that SPaTS consistently and significantly outperforms both frontier closed-source MLLMs and OCR MLLMs. Code is available at https://github.com/eeNickTang/SPaTS.

Rui Tang, Wentao Yang, Peirong Zhang +4

South China University of Technology · Guangzhou, China · HiThink Research +1

Visual Tokenization Scene Text Recognition Large Vision-Language Models Vision-Language Grounding RL for VLMs

Jul 30, 2026cs.CV

RefineSVG: Visual Feedback-Driven Reinforcement Learning for Image-to-SVG Generation

We propose RefineSVG, a single-step closed-loop visual feedback framework that enables multimodal large language models (MLLMs) to perform high-fidelity image-to-SVG generation through self-correction. Existing MLLM-based approaches rely on single-pass open-loop inference, where the model receives visual input only once and must generate thousands of SVG code tokens without intermediate verification. This paradigm inevitably leads to geometric drift, error accumulation, and visual hallucination on complex images. RefineSVG overcomes this limitation by invoking an external rendering engine after an initial SVG generation pass to compare the rendered output against the target image. The comparison yields a multi-dimensional visual residual map (Diff-Map) that is fed back to the model as a ReAct-style correction signal, driving a targeted correction step. To support this render-observe-correct interaction, we further introduce an SVG-oriented semantic vocabulary that compresses token sequences by over 52%. A progressive training pipeline spanning supervised fine-tuning, rejection-sampling cold-start data construction, and end-to-end agentic reinforcement learning aligns the model with closed-loop visual correction. Extensive experiments show that RefineSVG consistently outperforms existing baselines in reconstruction fidelity, structural accuracy, and code efficiency.Code is available at https://github.com/liuxiaobo66/RefineSVG.

Shaobo Liu, Feiqiao Mao, Shuaishuai Zhou +4

Shenzhen University · Shenzhen, China · Peking University +1

Image Generation Large Vision-Language Models RL for VLMs

Jul 29, 2026cs.RO

RL $^2$ -VLA: Adaptive RL Latent Compositional Steering with Test-Time Scaling for Vision-Language-Action Models

Despite the impressive visuomotor capabilities enabled by Vision-Language-Action (VLA) models, their performance often degrades on challenging and out-of-domain tasks. Recent test-time steering and scaling methods improve performance without extensive data collection and retraining, but action samples often remain concentrated around similar behaviors and therefore inherit correlated failure modes. Moreover, existing methods apply the same intervention strategy at every timestep, regardless of whether the base policy is already likely to succeed. To address these limitations, we introduce

RL^2

, an adaptive inference-time steering framework that leverages Reinforcement Learning on VLA Latents. First, we train a lightweight offline RL policy conditioned on expressive latents extracted from the VLA action expert and compose its flow velocity with that of the frozen VLA during inference. This compositional steering strategy combines the behavioral priors of large-scale imitation learning with the action diversity induced by offline RL beyond dominant demonstration modes. We further discover that inference-time steering follows fundamentally different scaling laws under success and failure states, revealing that action diversity is most beneficial when the base VLA is likely to fail, but can unnecessarily perturb already-accurate actions when success is likely. Building on this insight,

RL^2

activates compositional steering only when failure is predicted. Across the SIMPLER and PolaRiS benchmarks,

RL^2

improves success rates by up to +17.3% in out-of-domain settings, while ablations and scaling studies demonstrate the importance of latent representations and RL training. Finally, real-world experiments demonstrate that these gains transfer beyond simulation, establishing

RL^2

as a practical and modular steering framework for VLA deployment.

Derek Ming Siang Tan, Shailesh Shailesh, Srikrishna Iyer +4

1National University of Singapore · 3Singapore Technologies Engineering · University of Toronto

Vision-Language-Action Models Test-Time Adaptation of VLA Models Robustness of VLA Models RL for VLMs

Jul 28, 2026cs.AI

ODYSSE: Episode-wise Policy Optimization for Personalized Agentic Reasoning

Agentic systems have rapidly advanced in their ability to interact with real-world environments, leverage external tools, and provide services for users. However, unlike natural-world tasks that assume well-defined instructions, human-centered scenarios are characterized by ambiguous requests that lead to large, open-ended solution spaces. Decoding users' personalized preferences is therefore essential for narrowing the candidate solution space. This introduces a new challenge, personalized agentic reasoning, which requires agents to jointly interact with both users and environments to deliver personalized services. In this paper, we present ODYSSE, a Reinforced Fine-Tuning (RFT) framework for personalized agentic reasoning. At its core, ODYSSE proposes Episode-wise GRPO (ESPO), a novel extension of Group Relative Policy Optimization (GRPO) designed to address long action horizons and strong cross-step dependencies in personalized agentic reasoning. Rather than optimizing individual steps independently, ESPO introduces an episode-level reward mechanism together with episodic advantage estimation, enabling upstream evidence to effectively guide downstream personalized decisions and allowing agents to progressively resolve ambiguous user requests across multiple interaction steps. We further propose an episodic batch sampler that groups actions from the same episode into unified training batches, facilitating coherent optimization under ESPO. We evaluate ODYSSE on realistic long-horizon personalized GUI reasoning tasks. Experimental results demonstrate that ODYSSE consistently outperforms both specialist and general-purpose LVLMs, highlighting its effectiveness for personalized agentic reasoning.

Jiaqi Zhang, Tong Chen, Junliang Yu +2

The University of Queensland · Griffith University

Group Relative Policy Optimization Agentic Reasoning Long-Horizon Agent Tasks RL for VLMs

Jul 26, 2026cs.AI

Hybrid Advantage Estimation with Unified Critic for VLM Agentic Reinforcement Learning

Large Vision-Language Models (VLMs) now act as agents in interactive environments, where success requires coherent reasoning and decision-making across turns. Although end-to-end training in agentic environments can improve such multi-turn decision-making abilities, current methods mainly rely on either token-wise optimization over concatenated token trajectories or turn-wise optimization with uniform within-turn credit. In this work, we establish theoretical formulations for the two levels of optimization and derive a hybrid advantage that serves both objectives. Furthermore, with an appropriate choice of discount factor and learning target, we prove that a unified critic model can estimate values for both turn-wise and token-wise. As such, we propose HyGAE, an actor-critic framework that jointly optimizes token- and turn-level objectives with the hybrid advantage and unified critic. We conduct extensive evaluations of HyGAE across five multi-turn decision-making environments, where it achieves an average success rate of 91% and a significant improvement of 10% over other methods. Furthermore, we provide an in-depth analysis showing that the exact analytic form of the hybrid advantage and return is crucial for optimization. Project Page: https://wx-zhang.github.io/hygae-web/.

Wenxuan Zhang, Yuhui Wang, Donggang Jia +5

KAUST

Agentic RL Actor-Critic Methods RL for VLMs

Jul 23, 2026cs.CV

Be Consistent! Enhancing Robust Visual Reasoning in LVLMs with Consistency Constraints

While Large Vision-Language Models (LVLMs) exhibit strong perceptual capabilities, they remain vulnerable in visual reasoning tasks. Existing benchmarks largely focus on symbolic mathematical or scientific problems and simple vision-centric tasks, offering limited assessment of complex visual reasoning and logical consistency, a critical requirement for reliable reasoning systems. We introduce ConVBench, a complex vision-centric reasoning benchmark in which each image is paired with two logically equivalent questions across six categories: action and state, complex counting, spatial reasoning, causal and intent understanding, commonsense reasoning, and temporal perception. To complement this benchmark, we define two evaluation metrics, logical consistency and robust accuracy, that jointly assess both the correctness and consistency of model responses. We further present ConVLM, which improves LVLM reasoning through Group Relative Policy Optimization (GRPO)-based reinforcement learning with a novel consistency reward. This method leverages automatically generated logically equivalent question-answer pairs and a dual-reward design combining accuracy- and consistency-based signals, encouraging agreement between paired responses. The framework functions effectively with or without strict answer supervision.

Liqiang Jing, Xiong Zhou, Siddharth Varia +3

University of Texas at Dallas · Amazon Web Services

VLM Evaluation Visual Reasoning VLM Reasoning RL for VLMs

Jul 23, 2026cs.AI

MIRROR: Learning from the Other View for Multi-Modal Reasoning

Unlike large language models (LLMs) that exhibit strong reasoning capabilities, vision-language models (VLMs) struggle with visual reasoning, even on geometry problems that admit equivalent text, diagram, and combined diagram+text views. We show that these views often elicit different behaviors: a model may solve a problem from text but fail on the corresponding diagram, or succeed visually while failing textually. This inconsistency suggests that different views expose complementary reasoning paths and failure modes that standard multimodal post-training does not fully exploit. To study and exploit this phenomenon, we construct ODA-Data, a high-quality paired multimodal geometry dataset with text-dominant, image-dominant, and combined image+text views of the same problems, together with splits for training and evaluating modality-dependent reasoning behaviors. We then develop Modality-Informed Reciprocal Reasoning Optimization (MIRROR), a reinforcement learning approach for improving multimodal reasoning via self supervision. For each problem, MIRROR evaluates the model under all views, selects the best-performing view as a teacher, and trains other views with a reverse-KL objective towards the teacher. Across reasoning benchmarks that evaluate on geometry problems, MIRROR improves over standard RL and yields more accurate and consistent behavior across modalities

Wen Ye, Yuxiao Qu, Aviral Kumar +1

University of Southern California · Carnegie Mellon University

Cross-Modal Knowledge Distillation Multimodal Reasoning VLM Reasoning RL for VLMs

Jul 22, 2026cs.CV

Trace: A Taxonomy-Guided Environment for Multidomain Visual Reasoning

Reinforcement learning with verifiable rewards (RLVR) has substantially improved language-model reasoning, yet its extension to vision-language models remains constrained by the lack of training data that are simultaneously broad, exactly verifiable, and reproducible. We introduce Trace, a taxonomy-guided environment for multidomain visual reasoning. Trace factorizes task construction into a scene grammar and an executable task program, separating visual realization from answer computation. A shared semantic state determines the rendered image, prompt, typed answer, verifier state, and replayable instance trace. The resulting environment comprises 1,000 tasks over 277 scene grammars and 11 visual domains, with controlled semantic and visual variation. RLVR on 64,000 Trace instances improves the macro-average across 24 external benchmarks by 3.51 percentage points for Qwen2.5-VL-3B and 4.06 points for Qwen2.5-VL-7B, providing evidence that broad procedural training can transfer beyond the generated task distributions. Project page: https://maveryn.github.io/trace/.

Md Tanvirul Alam

Rochester Institute of Technology Rochester, NY, USA

Synthetic Data Generation VLM Reasoning Reinforcement Learning with Verifiable Rewards RL for VLMs

Jul 21, 2026cs.CV

DobicVLM: Aligning Chest X-Ray Report Generation with Clinically-Grounded Programmatic Rewards via Group Relative Policy Optimization

Medical imaging is a cornerstone of diagnostics, yet automated chest X-ray report generation struggles with structural adherence, anatomical completeness, and semantic faithfulness. We introduce DobicVLM, a vision-language model combining supervised fine-tuning on MedGemma-4B with Group Relative Policy Optimization (GRPO) and clinically-grounded programmatic rewards. Our approach uses interpretable, rule-based reward components; structural verification, anatomical checklist, semantic similarity, and length constraints to enforce clinical standards without neural reward models. Trained on 1,000 de-identified image-report pairs from a private clinical dataset (with ethics approval and compliance to local regulations), DobicVLM is evaluated via blinded expert review on 69 held-out cases. DobicVLM outperforms Gemini 2.5 Flash across the majority of criteria, achieving the highest impression accuracy (27.2%) and medical terminology (86.5%) compared to both Gemini 2.5 Flash and MedGemma 4B baselines, with minor trade-offs in completeness and referrals. This demonstrates GRPO's value for transparent alignment in resource-limited settings. Keywords: Vision-Language Models, Radiology Report Generation, Reinforcement Learning, Medical AI, GRPO

Thanni Adewuyi, Angelica Obayi, Andem Aniekan +10

Dobic Health · HelpMum Africa · University of Ibadan

Radiology Report Generation Radiology VLMs RL for VLMs

Jul 17, 2026cs.CV

WeedExpert-R1: Incentivizing Botanical Reasoning in MLLMs with Reinforcement Learning for Precision Weed Grounding

Precision weed control requires species-level identification and instance-level localization. However, conventional object detectors use a closed vocabulary, limiting their deployment across regions, and cannot explain their predictions in complex agricultural scenes. Multimodal large language models (MLLMs) offer visual grounding and reasoning capabilities, but insufficient botanical knowledge can cause hallucinations in fine-grained weed identification. This study introduces WeedExpert-R1, a multimodal model that learns visually grounded botanical reasoning through verifiable rewards. A domain-specific Chain-of-Thought synthesis pipeline combines a human-curated botanical trait dictionary with an Auditor-Synthesizer LLM workflow to generate reasoning data for supervised fine-tuning. Group Relative Policy Optimization is then applied with rewards for format, accuracy, instance count, and response length. Across 37 weed species from six datasets, WeedExpert-R1-4B achieved 75.82 percent exact-set precision at an IoU threshold of 0.5, 89.30 percent precision, and 87.81 percent recall. It outperformed proprietary models, including GPT-5.4 and Gemini-3.1-Pro, and larger open-source models, including Qwen3-VL-30B-Instruct and Gemma-4-31B-it. Results on unseen species further demonstrate its open-vocabulary capability and potential for deployment across diverse regions and crops without retraining.

Zonglin Yang, Wei-Zhen Liang, Nevin Lawrence +4

aDepartment of Biological Systems Engineering, University of Nebraska-Lincoln, Lincoln, NE, USA · bPanhandle Research and Extension Center, University of Nebraska-Lincoln, Scottsbluff, NE, USA · cDepartment of Agronomy and Horticulture, University of Nebraska-Lincoln, Lincoln, NE, USA +1

Open-Vocabulary Object Detection Weed Detection Multimodal Reasoning Vision-Language Grounding RL for VLMs

Jul 16, 2026cs.AI

Stop Thinking, Start Looking: Efficient Post-Training for Multimodal Document Question Answering via Reasoning-Free Alignment

Efficient multimodal document question answering with explicit visual grounding, locating the precise document region that supports each answer remains an open challenge. Current approaches bifurcate into Supervised Fine-Tuning (SFT), which requires large annotated datasets and reaches optimization plateaus, and reasoning-centric Reinforcement Learning (RL), which depends on verbose intermediate traces that inflate inference token cost without clear benefit. We introduce Perception-RFT, a training framework that applies Group Relative Policy Optimization (GRPO) to multimodal document QA, bypassing intermediate reasoning tokens to directly align visual features with structured grounding outputs. To rigorously evaluate the necessity of reasoning, we construct a reasoning variant under identical reward settings. We find that reasoning-enabled models suppress their reasoning traces during training, converging to direct perception-based policies at the 4B parameter scale, reducing per-query inference token length by more than 60%, while reasoning-enabled RL underperforms perception-only training. Through a fine-grained analysis of Qwen3-VL-4B optimization dynamics, we confirm that SFT saturation and cold-start RL instability established in text-domain post-training extend to multimodal, and identify a previously uncharacterized Grounding Divergence: a selective trade-off between semantic robustness and geometric precision on two out of distribution (OOD) benchmarks (4,828 samples) under joint RL optimization. We further show that an early SFT

\rightarrow

RL transition achieves comparable precision with 65% less training data.

Harikrishnan P M, Goutham Vignesh, Ganesh Parab +4

Phi Labs, Quantiphi, India · Phi Labs, Quantiphi, Canada.

Multimodal Grounding Efficient Multimodal Inference Document QA Multimodal Foundation Models RL for VLMs

Jul 15, 2026cs.CV

SD-MAR: Multi-image Analytical Reasoning via Synthetic Data and Reinforcement Learning

Vision Language Models (VLMs) demonstrate strong perceptual abilities but remain limited in tasks requiring analytical reasoning across multiple visual states, such as multi-image comparison, change detection, and multi-step visual inference. These capabilities are critical for real-world multimodal applications where reasoning must be grounded in systematic differences between visual contexts. However, existing benchmarks rarely require both explicit visual comparison and analytical reasoning, leaving this capability underexplored. To address this gap, we introduce SD-MAR (Synthetic Data for Multi-image Analytical Reasoning), a framework for training and evaluating VLMs on multi-image analytical reasoning. SD-MAR constructs paired visual scenarios through controlled perturbations and generates reasoning tasks spanning semantic change attribution and quantitative comparison. We further train VLMs using GRPO-lite with Backward Discounted Allocation (BDA), a reinforcement learning approach that removes KL regularization to encourage stronger policy optimization while allocating greater credit to the later reasoning steps where analytical conclusions are formed. Experiments on Qwen2.5-VL-7B and InternVL3-8B show that GRPO-lite fine-tuning on SD-MAR improves in-domain accuracy by up to 36.95%, with Qwen2.5-VL-7B outperforming GPT-4.1 on the SD-MAR benchmark. Importantly, out-of-domain generalization is preserved or improved: performance remains within 1% on MME, MMMU-Pro, and MathVista, while improving by up to 4% on MMBench. LLM-as-judge evaluation further demonstrates consistent improvements in logical coherence and explanation quality across both models.

Shiyu Yuan, Sourav Sanjukta Bhabesh, Zhe Wang +3

1Stevens Institute of Technology · 2AGI Foundations for AWS · 3Amazon Web Services (AWS), Amazon

Visual Reasoning Synthetic Data Generation Multimodal Reasoning VLM Reasoning RL for VLMs

Jul 15, 2026cs.CV

SIVA-RL: Sensitivity-Invariance Visual Alignment for Multimodal Reinforcement Learning

Reinforcement learning with verifiable rewards (RLVR) drives multimodal reasoning, but answer-level correctness does not guarantee that a vision-language model grounds its predictions in visual evidence. Existing visual-intervention methods contrast policy behavior on original and modified images, yet assign supervision by the type of intervention rather than its observed effect. This assumption fails: identical operators produce heterogeneous outcomes across samples. We propose SIVA-RL, a Sensitivity-Invariance Visual Alignment framework that replaces operator-conditioned regularization with sample-wise, outcome-conditioned supervision. SIVA-RL constructs localized interventions through token-aligned, distance-constrained within-image PatchSwap. A frozen audit policy then scores each clean-intervention pair, and the observed reward drop becomes soft routing weights. Large-drop pairs drive sensitivity alignment, low-drop pairs drive clean-anchored invariance alignment, and ambiguous pairs are down-weighted. This design decouples intervention construction from supervision assignment and is compatible with both GRPO and DAPO backbones. Across nine multimodal reasoning benchmarks spanning mathematical, logical, and vision-dependent tasks, SIVA-RL improves 3B and 7B models over matched RL baselines in every setting. It yields an 8.79 percentage-point gain on vision-dependent reasoning and up to 14.9% relative overall improvement across all four GRPO- and DAPO-based configurations.

Cheng Tang, Junzhi Ning, Min Cen +9

Shanghai Artificial Intelligence Laboratory · Shanghai Jiao Tong University · Sichuan University +2

Reinforcement Learning with Verifiable Rewards Visually Grounded Reasoning Vision-Language Alignment RL for VLMs

Jul 15, 2026cs.RO

Joint On-and-Off Policy Learning for Vision-and-Language Navigation

Vision-and-Language Navigation (VLN) necessitates an embodied agent to navigate in the physical world by adhering to natural language instructions. Recent advancements in Vision-Language Models (VLM) have propelled the development of VLM-based VLN methods with two predominant paradigms: (1) imitation learning (IL) on expert demonstrations, followed by the Dataset Aggregation (DAgger) algorithm to bolster error recovery capabilities; (2) reinforcement learning (RL) driven by verifiable rewards to enhance reasoning and exploration. A notable gap is the absence of integration between these two distinct paradigms. This paper introduces JOP-VLN, a novel VLN framework that synergistically combines off-policy imitation learning and on-policy exploration within a three-stage training pipeline. Initially, IL is employed on expert demonstrations to acquire basic navigation skills. Subsequently, the DAgger algorithm is utilized to generate heuristic exploration trajectories, which are then used for imitation learning to improve error recovery capabilities. Finally, a joint on-and-off policy learning framework is implemented, featuring high-entropy trajectory sampling to enhance RL training efficiency and an error-correction-prioritized trajectory sorting strategy for effective error correction. Extensive experiments demonstrate the efficacy of JOP-VLN, achieving success rates of 69.9% and 68.0% on the VLN-CE R2R and RxR benchmarks, respectively, setting a new state-of-the-art on R2R. Project page: https://qingrongh.github.io/JOP-VLN.

Qingrong He, Lin Zhao, Kevin Zheng +1

All authors are with Joy Future Academy.

Vision-Language Navigation Interactive Imitation Learning RL for VLMs

Jul 14, 2026cs.RO

ChunkFlow: Towards Continuity-Consistent Chunked Policy Learning

Vision-language action (VLA) models increasingly adopt chunked action heads to satisfy real-time constraints; however, this introduces boundary jitter: overlapping regions between consecutive chunks often yield inconsistent predictions, degrading temporal coherence and the task success rate. Existing methods, such as inference-time blending, merely reweight mismatched proposals without correcting underlying errors, leading to residual accumulation under biased or noisy histories. We propose ChunkFlow, a seam-aware training-and-execution framework for chunked policies that aligns chunk structure with boundary execution. It partitions each chunk into frozen, editable, and future zones, applies deterministic overlap blending at execution, and trains raw predictions with seam and first- and second-order continuity losses. History corruption and scheduled sampling improve robustness to executed-history errors, while an AWAC fine-tuning stage adapts the policy without removing these structural regularizers. Under mild smoothness assumptions, pre-blending seam discrepancies provably decay with increasing overlap. Experiments on CALVIN, LIBERO, and real robots show an improved success-stability trade-off with low-latency inference. Project page: https://cytoderm-ai.github.io/chunkflow.

Zhao Yang, Yinan Shi, Mingyuan Yao +3

State Key Laboratory of Human-Machine Hybrid Augmented Intelligence, Institute of Artificial Intelligence and Robotics, Xi’an Jiaotong University, China. · Jiangsu Cytoderm Intelligent Technology Co., Ltd., China.

Visuomotor Policy Learning Vision-Language-Action Models Robustness of VLA Models RL for VLMs

Jul 14, 2026cs.RO

ExToken: Structured Exploration for Efficient Vision-Language-Action Reinforcement Fine-tuning

Reinforcement Learning (RL) has demonstrated significant potential for improving Vision-Language-Action (VLA) models on complex manipulation tasks. However, its practical scalability remains severely limited by the substantial cost of environmental interactions. In this work, we first investigate the exploration stagnation bottleneck in current VLA-RL frameworks and reveal that trajectory diversity is fundamentally more important to sample efficiency than the sheer quantity of collected rollouts. Motivated by these insights, we introduce RL Exploration Token (ExToken), a simple yet general framework that condition VLA policies on discrete behavioral priors derived from offline demonstrations for structured exploration. By conditioning the policy on different tokens during rollout collection, ExToken encourages the agent to explore diverse behavioral modes, substantially improving state-action coverage and exploration efficiency. To bridge exploration during training with deterministic inference at deployment, ExToken further incorporates a state-conditioned token selector that adaptively predicts effective behavioral modes for unseen scenarios. Extensive experiments across simulated and real-world robotic manipulation tasks demonstrate that ExToken consistently accelerates convergence, improves task performance, and exhibits strong robustness under highly constrained interaction budgets.

Yilun Kong, Yunpeng Qing, Guozheng Ma +4

Nanyang Technological University · ACE Robotics · Zhejiang University +1

RL Exploration Robotic Manipulation Vision-Language-Action Models Sample-Efficient RL RL for VLMs

Jul 14, 2026cs.AI

A Learning-Rate-Gated Failure of GRPO in a Small Language and Vision-Language Model Web Agent: A Controlled Null and Its Mechanism

Reinforcement learning with verifiable rewards, and Group Relative Policy Optimization (GRPO) in particular, is now run routinely on a supervised checkpoint in the hope of producing a stronger agent. We ask whether it adds skill to a small language and vision-language model web agent at the 4B to 8B scale, or whether it mostly reshapes behavior the supervised model already has. Across a control grid of 18 runs that varies learning rate, KL weight, seed, initialization, and clipping, no configuration credibly improves the success rate of a strong supervised baseline on tasks the agent has largely mastered. On the text track, moderate to high learning rates make it credibly worse. The null holds under paired testing, 25 evaluation seeds, 6 training seeds, changes to the recipe, both text and Set-of-Marks screenshot observations, and scaling the backbone to 8B; the credible harm is a text-track finding and is only nominal under Set-of-Marks. To show that the null reflects the setting and not a broken pipeline, we run the identical harness, reward, and recipe on tasks whose reward is reachable by sampling, and there the success rate rises by 22 points with a paired interval that excludes zero. GRPO therefore helps only when there is headroom to climb, meaning the sampled policy already succeeds more often than the greedy one. We then explain the failure. A middle learning rate degrades the agent and a high one collapses it, and the two regimes form a double dissociation: grafting localizes the degrade regime to the attention and MLP blocks, while the collapse regime cannot be traced to any single group, and the embedding change that dominates the weight movement is causally inert. At 4B, effective rank in the late layers tracks capability in both directions; at 8B the two come apart. This coupling is specific to the smaller model, so we report it as scale-dependent.

Chengguang Gan, Zhixi Cai, Yunhao Liang +3

Independent Researcher · Monash University · University of Chinese Academy of Sciences +2

Group Relative Policy Optimization RL Fine-Tuning RL for VLMs RL for LLM Agents Web Agents

Jul 13, 2026cs.CV

Actor as Its Own Critic: Unifying Region Understanding and Localization via CycleGRPO

This paper introduces Actor as Its Own Critic, a unified reinforcement learning framework, Cycle Group Relative Policy Optimization (CycleGRPO), that jointly optimizes region understanding and localization for Multimodal Large Language Models (MLLMs). Unlike existing separate pipelines, we leverage the inherent duality between the two tasks to construct a self-evaluating reinforcement learning paradigm: "region

\to

text

\to

region''. Specifically, a single MLLM first acts as the actor to generate region captions, then immediately transitions to a critic to ground its generated text back in the spatial domain. Therefore, CycleGRPO requires only region inputs, e.g., masks or bounding boxes, entirely bypassing the need for textual ground truths. A quality-aware token-level cycle-consistency reward is employed to assess the semantic discriminability of text captions via their physical localization accuracy. Empirically, built upon SAMTok, our CycleGRPO framework successfully bootstraps both capabilities simultaneously. Without any task-specific fine-tuning, the framework yields consistent performance gains across a wide range of benchmarks, including region captioning, region VQA, grounded dialogue, and referring segmentation. Overall, CycleGRPO offers a straightforward and scalable way to advance pixel-level capabilities in MLLMs. Code and models are released at https://github.com/devinxzhang/CycleGRPO.

Xin Zhang, Haochen Wang, Yikang Zhou +3

National University of Singapore · ByteDance

Group Relative Policy Optimization Multimodal Large Language Models Referring Image Segmentation Vision-Language Grounding RL for VLMs

Jul 13, 2026cs.AI

SVR-R1: Bootstrapping Multi-modal Reasoning with Self-verification in Reinforcement Learning

We introduce Self-Verified Reasoner (SVR-R1), a multi-turn RL framework that turns a model's own verification into a learning signal for multimodal reasoning. For each query, the model proposes an answer using the same weights, and issues a binary self-verdict (Yes/No). A 'No' triggers a second-chance rethink; a 'Yes,' or a turn cap, finalizes the output for computing the outcome-based reward. SVR-R1 is implemented with GRPO and an asynchronous multi-turn rollout framework and needs no external supervision or auxiliary critics. We evaluate SVR-R1 on vision-language reasoning benchmarks and show that it improves accuracy by a large margin over strong standard GRPO baselines. Training dynamics show decreasing reliance on verification-fewer verification turns, yet higher test accuracy-indicating that the gap between verification and generation narrows as the policy internalizes self-correction and chooses the most confident answer via our framework. SVR-R1 bridges the less explored intersection of inference-time self-refinement and RL training for VLMs, offering a simple yet effective recipe for bootstrapping multimodal reasoning. We will open-source \textbf{SVR-R1} to facilitate future research in VLMs.

Mingyuan Wu, Jingcheng Yang, Shengyi Qian +11

University of Illinois Urbana-Champaign · Meta

Vision-Language Models Multimodal Reasoning VLM Reasoning RL for VLMs

Jul 11, 2026cs.CV

REVA-PO: Stabilizing Reinforcement Learning for Chest X-ray Report Generation

Automated chest X-ray report generation has recently benefited from reinforcement learning (RL) and large language models. However, RL training often suffers from instability or limited exploration due to fixed Kullback-Leibler (KL) regularization and a static reference policy that accumulates KL pressure over time. We propose Response-Weighted and Validation-Anchored Policy Optimization (REVA-PO), a RL framework that stabilizes long-term training via Response-Weighted Regularization (RER) and Validation-Anchored Policy Reset (VAPR). RER dynamically adjusts per-response KL weights based on advantage and reference-policy entropy, relaxing constraints for high-quality responses while tightening them for low-quality ones. Complementarily, VAPR periodically synchronizes the reference and current policies to the best validation checkpoint, resetting accumulated regularization pressure to expand the viable exploration space. To ensure a robust starting point, we employ a three-stage pipeline consisting of warm-up training, classifier-guided supervised fine-tuning, and RL. Extensive evaluations on MIMIC-CXR and IU-Xray demonstrate that REVA-PO sets new state-of-the-art benchmarks in both linguistic quality and clinical accuracy. Notably, BLEU-4 improves by 5.1% on MIMIC-CXR and 3.6% on IU-Xray, while CheXpert F1 and RadGraph F1 scores increase by 4.5% and 12.8%, respectively, over prior leading methods. The code is publicly available at https://github.com/LiGuo12/REVA_PO/.

Li Guo, Anas M. Tahir, Z. Jane Wang

University of British Columbia, Vancouver, BC, Canada

Radiology Report Generation Radiology VLMs KL-Regularized RL RL Post-Training RL for VLMs

Jul 10, 2026cs.AI

Multimodal Reward Hacking in Reinforcement Learning

Reinforcement learning (RL) is increasingly used to align multimodal large language models (MLLMs), but higher rewards do not always imply better task performance. This risk is amplified when visual evidence is evaluated by text-only or weakly grounded rewards. We study reward hacking in MLLM RL across safety VQA, chart VQA, and stress-test settings, varying reward design, data ambiguity, model scale (2B-32B), and RL algorithm (GRPO, RLOO, DAPO). We introduce Newly Rewarded Failure Rate (NRFR), which measures failures among samples whose proxy reward improves over the SFT baseline. Outcome-only rewards cause severe hacking, reaching 48.1% Reward Hacking Rate (RHR), while NRFR exceeding RHR shows that RL creates new failures rather than merely inheriting them. Scaling reduces but does not eliminate hacking: even the 32B model retains a 54.9% worse rate under outcome-only rewards, whereas answer-aware rewards improve the oracle trend at every scale. Robustness is also algorithm- and scale-dependent: GRPO is consistently most resistant, RLOO remains vulnerable, and DAPO improves substantially from 2B to 8B. Visual-evidence rewards help only with reliable verification: keyword-based checks increase hacking, while VLM-as-judge semantic verification reduces it. Overall, multimodal reward hacking is a systematic result of optimizing imperfect rewards, and robust alignment requires rewards and verifiers that remain reliable under optimization pressure.

Jiayu Yao, Yiwei Wang, Anmeng Zhang +5

Institute of Computing Technology, Chinese Academy of Sciences · University of California, Merced · Southeast University

Multimodal Large Language Models Multimodal Reward Modeling Reinforcement Learning with Verifiable Rewards RL for VLMs

Jul 10, 2026cs.LG

Learning from Hindsight for VLA Reinforcement Learning

Reinforcement learning is increasingly used to fine-tune vision-language-action (VLA) models, but robot interaction is expensive and learning becomes highly sample inefficient when successful rollouts are rare. When reward is assigned only for completing the commanded task, a failed rollout is treated as having no value even if it successfully executes behaviors relevant to that task. A robot that fails to place the correct object in a bowl may still move that object toward the bowl or place a different object inside it, demonstrating objects and actions that can be reused to solve the target task. These behaviors define auxiliary tasks that the policy can already solve, providing useful learning signals even before it can solve the harder target task. We introduce

\textit{Learning from Hindsight (LfH)}

, which turns such failures into additional learning signals. Using a pretrained vision-language model, LfH relabels failed rollouts with the behaviors they actually accomplish and trains the policy jointly on the commanded task and these auxiliary tasks. On out-of-distribution LIBERO-PRO manipulation tasks, LfH matches the final performance of GRPO with approximately

5\times

fewer rollouts and improves sample efficiency across multiple VLA backbones. On a physical Franka robot, LfH raises success from

0\%

to

56\%

within 160 training rollouts, while GRPO reaches

22\%

.

Iris Xu, Sunshine Jiang, John Marangola +2

Massachusetts Institute of Technology · MIT-IBM Computing Research Lab

Vision-Language-Action Models RL for VLMs