Group Relative Policy Optimization

The enhancement of reasoning capabilities in large language models (LLMs) has garnered significant attention, with supervised fine-tuning (SFT) and reinforcement learning emerging as dominant paradigms. While recent studies recognize the importance of reflection in reasoning processes, existing methodologies seldom address proactive reflection encouragement during training. This study focuses on mathematical reasoning by proposing a four-stage framework integrating Group Relative Policy Optimization (GRPO) with reflection reward mechanisms to strengthen LLMs' self-reflective capabilities. Besides, this approach incorporates established accuracy and format reward. Experimental results demonstrate GRPO's state-of-the-art performance through reflection-encouraged training, with ablation studies confirming the reflection reward's pivotal role. Comparative evaluations demonstrate full-parameter SFT's superiority over low-rank adaptation (LoRA) despite heightened computational demands. Building on these cumulative findings, this research substantiates GRPO's methodological significance in post-training optimization and envisions its potential to serve as a pivotal enabler for future LLM-based intelligent agents through the synergistic integration of cognitive rewards with dynamic environmental interactions.

Regulating the importance ratio is critical for the training stability of Group Relative Policy Optimization (GRPO) based frameworks. However, prevailing ratio control methods, such as hard clipping, suffer from non-differentiable boundaries and vanishing gradient regions, failing to maintain gradient fidelity. Furthermore, these methods lack a hazard-aware mechanism to adaptively suppress extreme deviations, leaving the optimization process vulnerable to abrupt policy shifts. To address these challenges, we propose Modulated Hazard-aware Policy Optimization (MHPO), a novel framework designed for robust and stable reinforcement learning. The proposed MHPO introduces a Log-Fidelity Modulator (LFM) to map unbounded importance ratios into a bounded, differentiable domain. This mechanism effectively prevents high-variance outlier tokens from destabilizing the loss landscape while ensuring global gradient stability. Complementarily, a Decoupled Hazard Penalty (DHP) integrates cumulative hazard functions from survival analysis to independently regulate positive and negative policy shifts. By shaping the optimization landscape with hazard-aware penalties, the proposed MHPO achieves fine-grained regulation of asymmetric policy shifts simultaneously mitigating mode collapse from over-expansion and preventing policy erosion from catastrophic contraction within a stabilized trust region. Extensive evaluations on diverse reasoning benchmarks across both text-based and vision-language tasks demonstrate that MHPO consistently outperforms existing methods, achieving superior performance while significantly enhancing training stability.

Group Relative Policy Optimization (GRPO) has emerged as an effective method for training reasoning models. While it computes advantages based on group mean, GRPO treats each output as an independent sample during the optimization and overlooks a vital structural signal: the natural contrast between correct and incorrect solutions within the same group, thus ignoring the rich, comparative data that could be leveraged by explicitly pitting successful reasoning traces against failed ones. To capitalize on this, we present a contrastive reformulation of GRPO, showing that the GRPO objective implicitly maximizes the margin between the policy ratios of correct and incorrect samples. Building on this insight, we propose Bilateral Context Conditioning (BICC), a mechanism that allows the model to cross-reference successful and failed reasoning traces during the optimization, enabling a direct information flow across samples. We further introduce Reward-Confidence Correction (RCC) to stabilize training by dynamically adjusts the advantage baseline in GRPO using reward-confidence covariance derived from the first-order approximation of the variance-minimizing estimator. Both mechanisms require no additional sampling or auxiliary models and can be adapted to all GRPO variants. Experiments on mathematical reasoning benchmarks demonstrate consistent improvements across comprehensive models and algorithms. Code is available at https://github.com/Skylanding/BiCC.

Group Relative Policy Optimization (GRPO) has emerged as a powerful framework for preference alignment in text-to-image (T2I) flow models. However, we observe that the standard paradigm where evaluating a group of generated samples against a single condition suffers from insufficient exploration of inter-sample relationships, constraining both alignment efficacy and performance ceilings. To address this sparse single-view evaluation scheme, we propose Multi-View GRPO (MV-GRPO), a novel approach that enhances relationship exploration by augmenting the condition space to create a dense multi-view reward mapping. Specifically, for a group of samples generated from one prompt, MV-GRPO leverages a flexible Condition Enhancer to generate semantically adjacent yet diverse captions. These captions enable multi-view advantage re-estimation, capturing diverse semantic attributes and providing richer optimization signals. By deriving the probability distribution of the original samples conditioned on these new captions, we can incorporate them into the training process without costly sample regeneration. Extensive experiments demonstrate that MV-GRPO achieves superior alignment performance over state-of-the-art methods.

Combining Chain-of-Thought (CoT) with Reinforcement Learning (RL) improves text-to-image (T2I) generation, yet the underlying interaction between CoT's exploration and RL's optimization remains unclear. We present a systematic entropy-based analysis that yields three key insights: (1) CoT expands the generative exploration space, while RL contracts it toward high-reward regions; (2) final reward is strongly negatively correlated with both the mean and variance of image-token entropy, highlighting the need to reduce uncertainty and instability; and (3) the entropy of the textual CoT directly governs downstream image quality, with lower-entropy CoTs leading to better generations. Motivated by these findings, we propose Entropy-Guided Group Relative Policy Optimization (EG-GRPO), a fine-tuning strategy that reallocates optimization budget by uncertainty: low-entropy tokens are excluded from reward-driven updates to preserve stability, while high-entropy tokens receive an entropy bonus that encourages structured exploration without collapse. Experiments on standard T2I benchmarks demonstrate that EG-GRPO achieves state-of-the-art performance.

Reinforcement Learning (RL) has become an effective paradigm for enhancing Large Language Models (LLMs) and visual generative models. However, its application in text-to-audio (TTA) generation remains largely under-explored. Prior work typically employs offline methods like Direct Preference Optimization (DPO) and leverages Contrastive Language-Audio Pretraining (CLAP) models as reward functions. In this study, we investigate the integration of online Group Relative Policy Optimization (GRPO) into TTA generation. We adapt the algorithm for Flow Matching-based audio models and demonstrate that online RL significantly outperforms its offline counterparts. Furthermore, we incorporate rewards derived from Large Audio Language Models (LALMs), which can provide fine-grained scoring signals that are better aligned with human perception. With only 470M parameters, our final model, Resonate, establishes a new SOTA on TTA-Bench in terms of both audio quality and semantic alignment.

Unified vision-language models have made significant progress in multimodal understanding and generation, yet they largely fall short in producing multimodal interleaved outputs, which is a crucial capability for tasks like visual storytelling and step-by-step visual reasoning. In this work, we propose a reinforcement learning-based post-training strategy to unlock this capability in existing unified models, without relying on large-scale multimodal interleaved datasets. We begin with a warm-up stage using a hybrid dataset comprising curated interleaved sequences and limited data for multimodal understanding and text-to-image generation, which exposes the model to interleaved generation patterns while preserving its pretrained capabilities. To further refine interleaved generation, we propose a unified policy optimization framework that extends Group Relative Policy Optimization (GRPO) to the multimodal setting. Our approach jointly models text and image generation within a single decoding trajectory and optimizes it with our novel hybrid rewards covering textual relevance, visual-text alignment, and structural fidelity. Additionally, we incorporate process-level rewards to provide step-wise guidance, enhancing training efficiency in complex multimodal tasks. Experiments on MMIE and InterleavedBench demonstrate that our approach significantly enhances the quality and coherence of multimodal interleaved generation.

Tool-based Agentic Reinforcement Learning (TARL) has emerged as a promising paradigm for training search agents to interact with external tools for a multi-turn information-seeking process autonomously. However, we identify a critical training instability that leads to catastrophic model collapse: Importance Sampling Distribution Drift(ISDD). In Group Relative Policy Optimization(GRPO), a widely adopted TARL algorithm, ISDD manifests as a precipitous decline in the importance sampling ratios, which nullifies gradient updates and triggers irreversible training failure. To address this, we propose Search Agent Policy Optimization (SAPO), which stabilizes training via a conditional token-level KL constraint. Unlike hard clipping, which ignores distributional divergence, SAPO selectively penalizes the KL divergence between the current and old policies. Crucially, this penalty is applied only to positive tokens with low probabilities where the policy has shifted excessively, thereby preventing distribution drift while preserving gradient flow. Remarkably, SAPO requires only one-line code modification to standard GRPO, ensuring immediate deployability. Extensive experiments across seven QA benchmarks demonstrate that SAPO achieves +10.6% absolute improvement (+31.5% relative) over Search-R1, yielding consistent gains across varying model scales (1.5B, 14B) and families (Qwen, LLaMA).

Modern code generation models exhibit longer outputs, accelerated capability growth, and changed training dynamics, rendering traditional training methodologies, algorithms, and datasets ineffective for improving their performance. To address these training bottlenecks, we propose MicroCoder-GRPO, an improved Group Relative Policy Optimization approach with three innovations: conditional truncation masking to improve long output potential while maintaining training stability, diversity-determined temperature selection to maintain and encourage output diversity, and removal of KL loss with high clipping ratios to facilitate solution diversity. MicroCoder-GRPO achieves up to 17.6% relative improvement over strong baselines on LiveCodeBench v6, with more pronounced gains under extended context evaluation. Additionally, we release MicroCoder-Dataset, a more challenging training corpus that achieves 3x larger performance gains than mainstream datasets on LiveCodeBench v6 within 300 training steps, and MicroCoder-Evaluator, a robust framework with approximately 25% improved evaluation accuracy and around 40% faster execution. Through comprehensive analysis across more than thirty controlled experiments, we reveal 34 training insights across seven main aspects, demonstrating that properly trained models can achieve competitive performance with larger counterparts.

Group Relative Policy Optimization (GRPO) is highly effective for post-training autoregressive (AR) language models, yet its direct application to diffusion large language models (dLLMs) often triggers reward collapse. We identify two sources of incompatibility. First, GRPO relies on importance ratios defined by sequence probabilities, which are intractable in dLLMs and must be estimated (e.g., via ELBO-based or mean-field likelihood proxies), yielding inherently noisy ratios. Second, standard GRPO's formulation is not designed for estimated ratios: its conditional clipping can be anomalously bypassed by model-agnostic estimation noise, producing gradient spikes, while its fixed group-size normalization amplifies gradient-magnitude fluctuations under high-variance ratio estimates. We show these effects form a self-reinforcing instability loop that drives policy drift and further increases ratio variance. To break this loop, we propose StableDRL, a reformulation of GRPO tailored for dLLMs that uses (i) unconditional clipping to suppress outlier-induced spikes and (ii) self-normalization to constrain updates within the convex hull of per-sample gradients. We further extend StableDRL to block-wise diffusion models via a staircase attention mechanism.

Group Relative Policy Optimization (GRPO) effectively scales LLM reasoning but incurs prohibitive computational costs due to its extensive group-based sampling requirement. While recent selective data utilization methods can mitigate this overhead, they could induce estimation bias by altering the underlying sampling distribution, compromising theoretical rigor and convergence behavior. To address this limitation, we propose Dynamic Pruning Policy Optimization (DPPO), a framework that enables dynamic pruning while preserving unbiased gradient estimation through importance sampling-based correction. By incorporating mathematically derived rescaling factors, DPPO significantly accelerates GRPO training without altering the optimization objective of the full-batch baseline. Furthermore, to mitigate the data sparsity induced by pruning, we introduce Dense Prompt Packing, a window-based greedy strategy that maximizes valid token density and hardware utilization. Extensive experiments demonstrate that DPPO consistently accelerates training across diverse models and benchmarks. For instance, on Qwen3-4B trained on MATH, DPPO achieves 2.37$\times$ training speedup and outperforms GRPO by 3.36% in average accuracy across six mathematical reasoning benchmarks.

Large Language Models remain vulnerable to adversarial prefix attacks (e.g., Sure, here is'') despite robust standard safety. We diagnose this vulnerability as Shallow Safety Alignment, stemming from a pathology we term semantic representation decay: as the model generates compliant prefixes, its internal malicious intent signal fades. To address this, we propose Two-Stage Causal-GRPO (TSC-GRPO), a framework designed to achieve intent pinning. First, grounded in causal identifiability theory, we train a causal intent probe to disentangle invariant intent from stylistic perturbations. Second, we internalize this causal awareness into the policy via Group Relative Policy Optimization. By employing a cumulative causal penalty within fork-in-the-road'' training scenarios, we force the model to learn that accumulating harmful tokens monotonically decreases reward, enabling robust late-stage refusals. Experiments show that TSC-GRPO significantly outperforms baselines in defending against jailbreak attacks while preserving general utility.

Text-to-audio (T2A) generation has advanced considerably in recent years, yet existing methods continue to face challenges in accurately rendering complex text prompts, particularly those involving intricate audio effects, and achieving precise text-audio alignment. While prior approaches have explored data augmentation, explicit timing conditioning, and reinforcement learning, overall synthesis quality remains constrained. In this work, we experiment with reinforcement learning to further enhance T2A generation quality, building on diffusion transformer (DiT)-based architectures. Our method first employs a large language model (LLM) to generate high-fidelity, richly detailed audio captions, substantially improving text-audio semantic alignment, especially for ambiguous or underspecified prompts. We then apply Group Relative Policy Optimization (GRPO), a recently introduced reinforcement learning algorithm, to fine-tune the T2A model. Through systematic experimentation with diverse reward functions (including CLAP, KL, FAD, and their combinations), we identify the key drivers of effective RL in audio synthesis and analyze how reward design impacts final audio quality. Experimental results demonstrate that GRPO-based fine-tuning yield substantial gains in synthesis fidelity and prompt adherence.

Group relative policy optimization (GRPO), a core methodological component of DeepSeekMath and DeepSeek-R1, has emerged as a cornerstone for scaling reasoning capabilities of large language models. Despite its widespread adoption and the proliferation of follow-up works, the theoretical properties of GRPO remain less studied. This paper provides a unified framework to understand GRPO through the lens of classical U-statistics. We demonstrate that the GRPO policy gradient is inherently a U-statistic, allowing us to characterize its mean squared error (MSE), derive the finite-sample error bound and asymptotic distribution of the suboptimality gap for its learned policy. Our findings reveal that GRPO is asymptotically equivalent to an oracle policy gradient algorithm -- one with access to a value function that quantifies the goodness of its learning policy at each training iteration -- and achieves asymptotically optimal performance within a broad class of policy gradient algorithms. Furthermore, we establish a universal scaling law that offers principled guidance for selecting the optimal group size. Empirical experiments further validate our theoretical findings, demonstrating that the optimal group size is universal, and verify the oracle property of GRPO.

Reinforcement learning (RL) with group relative policy optimization (GRPO) has become a widely adopted approach for enhancing the reasoning capabilities of multimodal large language models (MLLMs). While GRPO enables long-chain reasoning without a critic, it often suffers from sparse rewards on difficult problems and advantage vanishing when group-level rewards are too consistent for overly easy or hard problems. Existing solutions (sample expansion, selective utilization, and indirect reward design) often fail to maintain enough variance in within-group reward distributions to yield clear optimization signals. To address this, we propose DIVA-GRPO, a difficulty-adaptive variant advantage method that adjusts variant difficulty distributions from a global perspective. DIVA-GRPO dynamically assesses problem difficulty, samples variants with appropriate difficulty levels, and calculates advantages across local and global groups using difficulty-weighted and normalized scaling. This alleviates reward sparsity and advantage vanishing while improving training stability. Extensive experiments on six reasoning benchmarks demonstrate that DIVA-GRPO outperforms existing approaches in training efficiency and reasoning performance. Code: https://github.com/Siaaaaaa1/DIVA-GRPO

Distilling robust reasoning capabilities from large language models (LLMs) into smaller, computationally efficient student models remains an unresolved challenge. Despite recent advances, distilled models frequently suffer from superficial pattern memorization and subpar generalization. To overcome these limitations, we introduce a novel distillation framework that moves beyond simple mimicry to instill a deeper conceptual understanding. Our framework features two key innovations. First, to address pattern memorization, Explanatory Inversion (EI) generates targeted ``explanatory probes'' that compel the student to articulate the underlying logic behind an answer, rather than just memorizing it. Second, to improve generalization, Explanatory GRPO (EXGRPO) uses a reinforcement learning algorithm with a novel Dialogue Structure Utility Bonus, which explicitly rewards the student for maintaining a coherent reasoning process across these probes. Extensive evaluations on 12 datasets demonstrate significant improvements. Using Gemma-7b as the student model, our method yields an average 20.39% increase over zero-shot performance and a 6.02% improvement over the state-of-the-art distillation baselines. Moreover, models distilled with our method show remarkable training efficiency (e.g., surpassing vanilla fine-tuning with 10-25% training data) and strong generalization to out-of-distribution tasks. Implementation is released at https://github.com/Zhen-Tan-dmml/ExGRPO.git.

Group-based reinforcement learning (RL), such as GRPO, has advanced the capabilities of large language models on long-horizon agentic tasks. To enable more fine-grained policy updates, recent research has increasingly shifted toward stepwise group-based policy optimization, which treats each step in a rollout trajectory independently while using a memory module to retain historical context. However, we find a key issue in estimating stepwise relative advantages, namely context inconsistency, where steps within the same group may differ in their historical contexts. Empirically, we reveal that this issue can lead to severely biased advantage estimation, thereby degrading policy optimization significantly. To address the issue, in this paper, we propose Hierarchy-of-Groups Policy Optimization (HGPO) for long-horizon agentic tasks. Specifically, within a group of rollout trajectories, HGPO assigns each step to multiple hierarchical groups according to the consistency of historical contexts. Then, for each step, HGPO computes distinct advantages within each group and aggregates them with an adaptive weighting scheme. In this way, HGPO can achieve a favorable bias-variance trade-off in stepwise advantage estimation, without extra models or rollouts. Evaluations on two challenging agentic tasks, ALFWorld and WebShop with Qwen2.5-1.5B-Instruct and Qwen2.5-7B-Instruct, show that HGPO significantly outperforms existing agentic RL methods under the same computational constraints. Code is available at https://github.com/langfengQ/verl-agent/tree/master/recipe/hgpo.

Reinforcement Learning with Verifiable Rewards (RLVR) and Group Relative Policy Optimization (GRPO) have significantly advanced the reasoning capabilities of large language models. Extending these methods to multimodal settings, however, faces a critical challenge: the instability of std-based normalization, which is easily distorted by extreme samples with nearly positive or negative rewards. Unlike pure-text LLMs, multimodal models are particularly sensitive to such distortions, as both perceptual and reasoning errors influence their responses. To address this, we characterize each sample by its difficulty, defined through perceptual complexity (measured via visual entropy) and reasoning uncertainty (captured by model confidence). Building on this characterization, we propose difficulty-aware group normalization (Durian), which re-groups samples by difficulty levels and shares the std within each group. Our approach preserves GRPO's intra-group distinctions while eliminating sensitivity to extreme cases, yielding significant performance gains across multiple multimodal reasoning benchmarks.

Following the success of Group Relative Policy Optimization (GRPO) in foundation LLMs, an increasing number of works have sought to adapt GRPO to Visual Large Language Models (VLLMs) for visual perception tasks (e.g., detection and segmentation). However, much of this line of research rests on a long-standing yet unexamined assumption: training paradigms developed for language reasoning can be transferred seamlessly to visual perception. Our experiments show that this assumption is not valid, revealing intrinsic differences between reasoning-oriented and perception-oriented settings. Using reasoning segmentation as a representative case, we surface two overlooked factors: (i) the need for a broader output space, and (ii) the importance of fine-grained, stable rewards. Building on these observations, we propose Dr.~Seg, a simple, plug-and-play GRPO-based framework consisting of a Look-to-Confirm mechanism and a Distribution-Ranked Reward module, requiring no architectural modifications and integrating seamlessly with existing GRPO-based VLLMs. Extensive experiments demonstrate that Dr.~Seg improves performance in complex visual scenarios while maintaining strong generalization. Code, models, and datasets are available at https://github.com/eVI-group-SCU/Dr-Seg.

Vision-Language-Action (VLA) models are advancing autonomous driving by replacing modular pipelines with unified end-to-end architectures. However, current VLAs face two expensive requirements: (1) massive dataset collection, and (2) dense reasoning annotations. In this work, we address both challenges with NORD (No Reasoning for Driving). Compared to existing VLAs, NORD achieves competitive performance while being fine-tuned on <60% of the data and no reasoning annotations, resulting in 3x fewer tokens. We identify that standard Group Relative Policy Optimization (GRPO) fails to yield significant improvements when applied to policies trained on such small, reasoning-free datasets. We show that this limitation stems from difficulty bias, which disproportionately penalizes reward signals from scenarios that produce high-variance rollouts within GRPO. NORD overcomes this by incorporating Dr. GRPO, a recent algorithm designed to mitigate difficulty bias in LLMs. As a result, NORD achieves competitive performance on Waymo and NAVSIM with a fraction of the training data and no reasoning overhead, enabling more efficient autonomous systems. Website: https://nord-vla-ai.github.io/

Existing alignment paradigms remain limited in capturing the pluralistic nature of human values. Overton Pluralism addresses this gap by generating responses with diverse perspectives from a single query. This paper introduces OP-GRPO (Overton Pluralistic Group Relative Policy Optimization), a reinforcement learning framework for implicit Overton Pluralism that enables a single large language model to produce pluralistic responses without explicit prompting or modular orchestration. Our workflow consists of two main steps. First, similarity estimator training fine-tunes a Sentence Transformer for Overton Pluralism tasks to provide more accurate coverage evaluation of generated responses. Second, OP-GRPO training incorporates this similarity estimator into a dual-reward system designed to ensure both broad coverage of genuine human perspectives and the uniqueness of each perspective, thereby promoting diversity. Empirical results demonstrate a "small models, big perspective coverage" effect. The trained Qwen2.5-3B-Instruct model surpasses a 20B GPT-OSS baseline with a 37.4 percent relative accuracy gain on a Natural Language Inference benchmark, and also outperforms a modular architecture baseline with a 19.1 percent relative improvement. Additional evaluations using GPT-4.1 as a large language model judge further confirm the robustness of the approach.

Group Relative Policy Optimization (GRPO) has significantly advanced the training of large language models and enhanced their reasoning capabilities, while it remains susceptible to instability due to the use of hard clipping. Soft Adaptive Policy Optimization (SAPO) addresses this limitation by replacing clipping with a smooth sigmoid-based gate function, which leads to more stable updates. We have decided to push this theory further and investigate the impact of different gate functions on both training stability and final model performance. We formalize the key properties that admissible gates should satisfy and identify several families of such functions for empirical evaluation. This paper presents an analysis of our findings based on experiments conducted with the Qwen2.5-7B-Instruct model on mathematical reasoning tasks. These results provide practical guidance for designing smoother and more robust policy optimization objectives for large language model training.

Multi-turn LLM agents are becoming pivotal to production systems, spanning customer service automation, e-commerce assistance, and interactive task management, where accurately distinguishing high-value informative signals from stochastic noise is critical for sample-efficient training. In real-world scenarios, a failure in a trivial task may reflect random instability, whereas success in a high-difficulty task signifies a genuine capability breakthrough. Yet, existing group-based policy optimization methods rigidly rely on statistical deviation within discrete batches, frequently misallocating credit when task difficulty fluctuates. To address this issue, we propose Proximity-based Multi-turn Optimization (ProxMO), a practical and robust framework engineered specifically for the constraints of real-world deployment. ProxMO integrates global context via two lightweight mechanisms: success-rate-aware modulation dynamically adapts gradient intensity based on episode-level difficulty, while proximity-based soft aggregation derives baselines through continuous semantic weighting at the step level. Extensive evaluations on ALFWorld and WebShop benchmarks demonstrate that ProxMO yields substantial performance gains over existing baselines with negligible computational cost. Ablation studies further validate the independent and synergistic efficacy of both mechanisms. Crucially, ProxMO offers plug-and-play compatibility with standard GRPO frameworks, facilitating immediate, low-friction adoption in existing industrial training pipelines. Our implementation is available at: https://anonymous.4open.science/r/proxmo.

Large-scale flow matching models have achieved strong performance across generative tasks such as text-to-image, video, 3D, and speech synthesis. However, aligning their outputs with human preferences and task-specific objectives remains challenging. Flow-GRPO extends Group Relative Policy Optimization (GRPO) to generation models, enabling stable reinforcement learning alignment for generative systems. Since its introduction, Flow-GRPO has triggered rapid research growth, spanning methodological refinements and diverse application domains. This survey provides a comprehensive review of Flow-GRPO and its subsequent developments. We organize existing work along two primary dimensions. First, we analyze methodological advances beyond the original framework, including reward signal design, credit assignment, sampling efficiency, diversity preservation, reward hacking mitigation, and reward model construction. Second, we examine extensions of GRPO-based alignment across generative paradigms and modalities, including text-to-image, video generation, image editing, speech and audio, 3D modeling, embodied vision-language-action systems, unified multimodal models, autoregressive and masked diffusion models, and restoration tasks. By synthesizing theoretical insights and practical adaptations, this survey highlights Flow-GRPO as a general alignment framework for modern generative models and outlines key open challenges for scalable and robust reinforcement-based generation.

Group Relative Policy Optimization (GRPO) is effective for training language models on complex reasoning. However, since the objective is defined relative to a group of sampled trajectories, extended deliberation can create more chances to realize relative gains, leading to inefficient reasoning and overthinking, and complicating the trade-off between correctness and rollout efficiency. Controlling this behavior is difficult in practice, considering (i) Length penalties are hard to calibrate because longer rollouts may reflect harder problems that require longer reasoning, penalizing tokens risks truncating useful reasoning along with redundant continuation; and (ii) supervision that directly indicates when to continue or stop is typically unavailable beyond final answer correctness. We propose Weakly Supervised GRPO (WS-GRPO), which improves rollout efficiency by converting terminal rewards into correctness-aware guidance over partial trajectories. Unlike global length penalties that are hard to calibrate, WS-GRPO trains a preference model from outcome-only correctness to produce prefix-level signals that indicate when additional continuation is beneficial. Thus, WS-GRPO supplies outcome-derived continue/stop guidance, reducing redundant deliberation while maintaining accuracy. We provide theoretical results and empirically show on reasoning benchmarks that WS-GRPO substantially reduces rollout length while remaining competitive with GRPO baselines.

Despite their sophisticated general-purpose capabilities, Large Language Models (LLMs) often fail to align with diverse individual preferences because standard post-training methods, like Reinforcement Learning with Human Feedback (RLHF), optimize for a single, global objective. While Group Relative Policy Optimization (GRPO) is a widely adopted on-policy reinforcement learning framework, its group-based normalization implicitly assumes that all samples are exchangeable, inheriting this limitation in personalized settings. This assumption conflates distinct user reward distributions and systematically biases learning toward dominant preferences while suppressing minority signals. To address this, we introduce Personalized GRPO (P-GRPO), a novel alignment framework that decouples advantage estimation from immediate batch statistics. By normalizing advantages against preference-group-specific reward histories rather than the concurrent generation group, P-GRPO preserves the contrastive signal necessary for learning distinct preferences. We evaluate P-GRPO across diverse tasks and find that it consistently achieves faster convergence and higher rewards than standard GRPO, thereby enhancing its ability to recover and align with heterogeneous preference signals. Our results demonstrate that accounting for reward heterogeneity at the optimization level is essential for building models that faithfully align with diverse human preferences without sacrificing general capabilities.

Multi-agent LLM systems enable advanced reasoning and tool use via role specialization, yet reliable reinforcement learning (RL) post-training for such systems remains difficult. In this work, we theoretically pinpoint a key reason for training instability when extending group-based RL to multi-agent LLM systems. We show that under GRPO-style optimization, a global normalization baseline may deviate from diverse agents' reward distributions, which ultimately leads to gradient-norm instability. Based on this finding, we propose Dr. MAS, a simple and stable RL training recipe for multi-agent LLM systems. Dr. MAS uses an agent-wise remedy: normalizing advantages per agent using each agent's own reward statistics, which calibrates gradient scales and dramatically stabilizes training, both theoretically and empirically. Beyond the algorithm, Dr. MAS provides an end-to-end RL training framework for multi-agent LLM systems, supporting scalable orchestration, flexible per-agent LLM serving and optimization configs, and shared resource scheduling of LLM actor backends. We evaluate Dr. MAS on multi-agent math reasoning and multi-turn search benchmarks using Qwen2.5 and Qwen3 series models. Dr. MAS achieves clear gains over vanilla GRPO (e.g., +5.6% avg@16 and +4.6% pass@16 on math, and +15.2% avg@16 and +13.1% pass@16 on search) while largely eliminating gradient spikes. Moreover, it remains highly effective under heterogeneous agent-model assignments while improving efficiency.

Reinforcement Learning with Verifiable Rewards (RLVR) is commonly based on group sampling to estimate advantages and stabilize policy updates. In practice, computational limits often rule out very large groups, so training proceeds with finite rollout sets that can reinforce only the correct behavior they expose. At practical group sizes, updates can miss rare-correct trajectories while still containing mixed rewards, concentrating probability on more common sampled solutions. We derive the probability of such prompt-local tail-miss events as a function of group size, showing non-monotonic behavior, and in the categorical abstraction characterize how unsampled-correct mass can shrink even as total correct mass grows. Motivated by this analysis, we propose a difficulty-aware scaling coefficient, inspired by Focal loss, that down-weights updates on high-success sampled groups. Empirically, categorical simulation illustrates the same effect in the categorical setting, Maze provides a single-solution test, and LLM experiments include a representative GRPO group-size sweep together with fixed-$N$ transfer across GRPO, DAPO, and CISPO. On Qwen2.5-7B at $N{=}8$, our method improves average math pass@256 from 64.1 $\rightarrow$ 70.3 (GRPO), 69.3 $\rightarrow$ 72.5 (DAPO), and 73.2 $\rightarrow$ 76.8 (CISPO); OOD pass@256 also improves in all three cases, without increasing group size or computational cost.

Group Relative Policy Optimization (GRPO) remains the dominant critic-free approach for fine-tuning LLMs and VLMs, but its compatibility with constrained policy optimization (e.g. for safety-critical domains) has not been carefully examined. In this work, we introduce Constrained GRPO, a Lagrangian-based extension of GRPO for constrained policy optimization. We show that the standard practice of scalarizing rewards before normalization introduces a critical Lagrangian-specific failure mode: GRPO's within-group normalization makes constrained optimization highly sensitive to how multi-component learning signals are aggregated. We show that scalarizing rewards before normalization introduces shared-denominator coupling, so that changing one multiplier alters not only the emphasis on its corresponding constraint, but also the relative weighting of the reward and other constraints. We address this with a simple but crucial modification: scalarizing standardized advantages rather than rewards. This yields a better-conditioned update by addressing the coupling induced by reward scalarization, resulting in better-behaved multiplier dynamics and more stable constraint enforcement in practice. Empirically, across a controlled gridworld, a real-world autonomous driving benchmark, and a mathematical reasoning task, Constrained GRPO consistently achieves better adherence to specified constraints while maintaining or improving task performance.

Reinforcement Learning with Verifiable Rewards (RLVR), particularly GRPO, has become the standard for eliciting LLM reasoning. However, its efficiency in exploration and difficulty adaptation remains an open challenge. In this work, we identify an implicit advantage symmetry inherent in Group Relative Advantage Estimation (GRAE) as a structural property that provides a new perspective for understanding these bottlenecks. This symmetry induces two critical limitations: (i) at the group level, strict symmetry in weights between correct and incorrect trajectories leaves unsampled action logits unchanged, thereby hindering the exploration of novel correct solution. (ii) at the sample level, the algorithm implicitly prioritizes medium-difficulty samples, remaining agnostic to the non-stationary demands of difficulty focus. Through controlled experiments, we reveal that this symmetric property is sub-optimal, yielding two pivotal insights: (i) asymmetrically down-weighting the advantages of correct trajectories encourages essential exploration but risks instability; (ii) learning efficiency can be boosted by a curriculum-like transition-prioritizing simpler samples initially before gradually shifting to complex ones. Motivated by these findings, we propose Asymmetric GRAE (A-GRAE), which dynamically modulates exploration incentives and sample-difficulty focus. Experiments across seven benchmarks demonstrate that A-GRAE consistently improves GRPO and its variants across both LLMs and MLLMs. Code is available at https://github.com/HKU-HealthAI/A-GRAE