Sparse-Reward RL

Deterministic policy gradient (DPG) is widely utilized for continuous control; however, it inherently relies on the differentiability of the critic with respect to the action during policy updates. This assumption is violated in practical control problems involving sparse or discrete rewards, leading to ill-defined policy gradients and unstable learning. To address these challenges, we propose a principled alternative based on a smoothed Bellman equation formulated via Gaussian smoothing. Specifically, we define a novel action-value function based on a smoothed Bellman equation and derive the soft deterministic policy gradient (Soft-DPG). Our formulation eliminates explicit dependence on critic action-gradients and ensures that the gradient remains well-defined even for non-smooth Q-functions. We instantiate this framework into a deep reinforcement learning algorithm, which we call soft deep deterministic policy gradient (Soft DDPG). Empirical evaluations on standard continuous control benchmarks and their discretized-reward variants show that Soft DDPG remains competitive in dense-reward settings and provides clear gains in most discretized-reward environments, where standard DDPG is more sensitive to irregular critic landscapes.

Offline goal-conditioned RL (GCRL) learns goal-reaching policies from static datasets, but real-world datasets are often partially observable and history-dependent, exhibiting a mix of Markovian and non-Markovian that violate standard RL assumptions. History-aware sequence models such as Decision Transformer (DT) are a natural fit for long-term dependency modeling, yet pure attention is inefficient and brittle when handling local Markovian structure and long-range context simultaneously. Although recent hybrid architectures (e.g., LSDT) introduce local extractors to improve local dependencies modeling, the fixed-window extraction cannot adapt its effective memory to varying dependency lengths in temporally heterogeneous settings, often truncating long-range context rather than compressing its content adaptively. Moreover, sequential offline GCRL faces a key bottleneck: under sparse rewards, return-to-go (RTG) becomes non-discriminative across sub-trajectories, providing little guidance signal for stitching goal-reaching behaviors from diverse demonstrations. To address these, we propose QHyer, which replaces RTG with a flow-parameterized, state-conditioned goal-reaching Q-estimator to support stitching across demonstrations, and introduces a gated Hybrid Attention-Mamba backbone that performs content-adaptive history compression while preserving local dynamics. Extensive experiments demonstrate that QHyer achieves state-of-the-art performance on both non-Markovian and Markovian datasets, validating its effectiveness for diverse scenarios.

Artificial intelligence-generated content (AIGC) has emerged as a transformative paradigm for automating the creation of diverse and customized content, giving rise to rapidly growing computational workloads in cloud data centers. It is imperative for AIGC service providers (ASPs) to strategically schedule AIGC workloads to reduce data center energy costs while guaranteeing high-quality content generation. However, the distinctive characteristics of AIGC services pose critical challenges, including model heterogeneity across ASPs, implicit service quality evaluation, and complex inference process control. To tackle these challenges, we propose a joint energy management and coordinated AIGC workload scheduling framework, which introduces an explicit mathematical characterization of service quality to promote both job transfer among ASPs and fine-grained inference process configuration. Moreover, various energy resources within data centers are jointly considered to enhance power usage flexibility. Subsequently, a system utility maximization problem is formulated to balance AIGC service revenue with operational penalties and costs. Nevertheless, the strong coupling among job scheduling decisions induces severe reward sparsity, which limits the effectiveness of existing deep reinforcement learning (DRL) algorithms. To address this issue, we develop a diffusion model-aided reward shaping approach to synthesize complementary reward signals through a multi-step denoising process. This approach is seamlessly integrated with DRL to enable efficient learning of scheduling policies under sparse environmental feedback. Experiments based on real-world models and datasets demonstrate that our scheme effectively accommodates electricity price fluctuations and AIGC model heterogeneity, while achieving superior learning convergence and system utility compared with benchmark methods.

Reinforcement learning (RL) post-training has enabled newer capabilities in models, such as agentic tool-use for search. However, these models struggle primarily due to limitations with sparse outcome-based rewards and a lack of training data that encapsulates questions of differing hardness, which results in models not performing deeper searches with tools to collect evidence for question-answering. To address these limitations, we introduce S^3-R1 (Synthetic data and stabilized Search R1), a framework that couples a data-centric approach with denser learning signals. We first develop a synthetic generation and curation pipeline that programmatically derives diverse, multi-hop questions from existing documents. This pipeline incorporates a retrieval-based verification step to specifically isolate questions of intermediate difficulty. We then pair this expanded training set with a reward structure that evaluates both intermediate search quality and the correctness of the final answer. This setup directly mitigates the credit assignment problems inherent to sparse rewards. Our evaluations show that S^3-R1 outperforms existing baselines by learning more effective search and synthesis strategies, yielding up to a 10% improvement in robust generalization on out-of-domain datasets.

Reinforcement learning (RL) has substantially improved the ability of large language model (LLM) agents to interact with environments and solve multi-turn tasks. However, effective agentic RL remains challenging: sparse outcome-only rewards provide limited guidance for assigning credit to individual steps within long interaction trajectories. Existing approaches often introduce dense intermediate supervision, such as process reward models or auxiliary self-supervised signals, which increases supervision and tuning complexity and may limit generalization across tasks and domains. We present AEM, a supervision-free credit assignment method that adaptively modulates entropy dynamics during RL training to improve the exploration-exploitation trade-off. Since in agentic RL the environment is typically affected by a complete response, rather than an individual token, our analysis lifts entropy dynamics from the token level to the response level, aligning uncertainty estimation with the effective action granularity of LLM agents and reducing sensitivity to token-level sampling noise. We further show that entropy drift under natural-gradient updates is governed by the interaction between the sampled-response advantage and its relative surprisal. Motivated by this result, AEM derives a practical response-level uncertainty proxy and uses it to rescale advantages, leveraging the evolving balance between positive and negative samples to naturally transition from exploration to exploitation. Extensive experiments on ALFWorld, WebShop, and SWE-bench-Verified with models ranging from 1.5B to 32B demonstrate that AEM consistently improves strong RL baselines, including a +1.4% gain when integrated into a state-of-the-art software-engineering RL training framework.

Deep Reinforcement Learning (DRL) algorithms often require a large amount of data and struggle in sparse-reward domains with long planning horizons and multiple sub-goals. In this paper, we propose a neuro-symbolic extension of Proximal Policy Optimization (PPO) that transfers partial logical policy specifications learned in easier instances to guide learning in more challenging settings. We introduce two integrations of symbolic guidance: (i) H-PPO-Product, which biases the action distribution at sampling time, and (ii) H-PPO-SymLoss, which augments the PPO loss with a symbolic regularization term. We evaluate our methods on three benchmarks (OfficeWorld, WaterWorld, and DoorKey), showing consistently faster learning and higher return at convergence than PPO and a Reward Machine baseline, also under imperfect symbolic knowledge.

Many Multi-Agent Reinforcement Learning (MARL) agents fail to adapt properly to cooperating with agents trained with the same objectives but different seeds, algorithms, or other training differences. This is the problem of Zero-Shot Coordination (ZSC), which focuses on training agents to cooperate well with unknown agents. ZSC has been studied for a variety of tabular cases and simple games such as Hanabi, achieving excellent results. However, existing solutions to ZSC only consider identical rewards for your trained agents and all future partners. This is not realistic for the trained agents, as they do not consider the problem of cooperating with agents that have identical sparse objectives but shape the rewards for those objectives in different manner. To address this issue, we show how to train an ensemble of methods using randomized reward shapings chosen using 4 selection algorithms. Experiments done on the Overcooked environment demonstrate consistent improvements of 62.2%-119.2% in sparse reward over baseline ZSC algorithms when playing with agents that have identical sparse rewards but different reward shapings.

Generic group-based RL assumes that sampled rollout groups are already usable learning signals. We show that this assumption breaks down in sparse-hit generative recommendation, where many sampled groups never become learnable at all. We propose ReCast, a repair-then-contrast learning-signal framework that first restores minimal learnability for all-zero groups and then replaces full-group reward normalization with a boundary-focused contrastive update on the strongest positive and the hardest negative. ReCast leaves the outer RL framework unchanged, modifies only within-group signal construction, and partially decouples rollout search width from actor-side update width. Across multiple generative recommendation tasks, ReCast consistently outperforms OpenOneRec-RL, achieving up to 36.6% relative improvement in Pass@1. Its matched-budget advantage is substantially larger: ReCast reaches the baseline's target performance with only 4.1% of the rollout budget, and this advantage widens with model scale. The same design also yields direct system-level gains, reducing actor-side update time by 16.60x, lowering peak allocated memory by 16.5%, and improving actor MFU by 14.2%. Mechanism analysis shows that ReCast mitigates the persistent all-zero / single-hit regime, restores learnability when natural positives are scarce, and converts otherwise wasted rollout budget into more stable policy updates. These results suggest that, for generative recommendation, the decisive RL problem is not only how to assign rewards, but how to construct learnable optimization events from sparse, structured supervision.

The temporal lag between actions and their long-term consequences makes credit assignment a challenge when learning goal-directed behaviors from data. Generative world models capture the distribution of future states an agent may visit, indicating that they have captured temporal information. How can that temporal information be extracted to perform credit assignment? In this paper, we formalize how the temporal information stored in world models encodes the underlying geometry of the world. Leveraging optimal transport, we extract this geometry from a learned model of the occupancy measure into a reward function that captures goal-reaching information. Our resulting method, Occupancy Reward Shaping, largely mitigates the problem of credit assignment in sparse reward settings. ORS provably does not alter the optimal policy, yet empirically improves performance by 2.2x across 13 diverse long-horizon locomotion and manipulation tasks. Moreover, we demonstrate the effectiveness of ORS in the real world for controlling nuclear fusion on 3 Tokamak control tasks. Code: https://github.com/aravindvenu7/occupancy_reward_shaping; Website: https://aravindvenu7.github.io/website/ors/

Reinforcement learning for multi-step reasoning with large language models (LLMs) typically relies on sparse terminal rewards, which creates a poorly conditioned credit-assignment problem: the final feedback is propagated uniformly across all intermediate decisions. This leads to high gradient variance, unstable training, and many ineffective updates, ultimately limiting sustained model improvement. We propose a counterfactual-comparison framework for credit assignment. For each input, the framework samples multiple reasoning trajectories and treats their differences as implicit approximations to alternative decisions. This yields an implicit process-level advantage estimator that converts sparse terminal rewards into step-sensitive learning signals. Building on this framework, we introduce Implicit Behavior Policy Optimization (IBPO), which substantially improves training stability and the performance ceiling on mathematical and code-reasoning benchmarks. Our results point to a promising direction for unlocking the reasoning potential of LLMs.

Search agents extend Large Language Models (LLMs) beyond static parametric knowledge by enabling access to up-to-date and long-tail information unavailable during pretraining. While reinforcement learning has been widely adopted for training such agents, existing approaches face key limitations: process supervision often suffers from unstable value estimation, whereas outcome supervision struggles with credit assignment due to sparse, trajectory-level rewards. To bridge this gap, we propose Contribution-Weighted GRPO (CW-GRPO), a framework that integrates process supervision into group relative policy optimization. Instead of directly optimizing process rewards, CW-GRPO employs an LLM judge to assess the retrieval utility and reasoning correctness at each search round, producing per-round contribution scores. These scores are used to rescale outcome-based advantages along the trajectory, enabling fine-grained credit assignment without sacrificing optimization stability. Experiments on multiple knowledge-intensive benchmarks show that CW-GRPO outperforms standard GRPO by 5.0% on Qwen3-8B and 6.3% on Qwen3-1.7B, leading to more effective search behaviors. Additional analysis reveals that successful trajectories exhibit concentrated contributions in specific rounds, providing empirical insight into search agent tasks.

Group Relative Policy Optimization (GRPO) has significantly advanced the reasoning ability of large language models (LLMs), particularly in their mathemat ical reasoning performance. However, GRPO and related entropy regularization methods still struggle with token-level sparse-rewards, which is an inherent chal lenge in chain-of-thought (CoT) reasoning. These approaches often rely on undifferen tiated token-level entropy regularization, which easily leads to entropy collapse or model degradation under sparse token rewards. In this work, we propose TEPO, a novel token-level framework that (1) leverages sequence-level likelihood to link group-level rewards with individual tokens via token-level aggregation, and (2) introduces a token-level KL-Divergence mask constraint that targets tokens with positive advantages and decreasing entropy to mitigate abrupt policy updates. Experiments demonstrate that TEPO not only achieves state-of-the-art performance on mathematical reasoning benchmarks but also markedly enhances training stability, reducing convergence time by 50% compared with GRPO/DAPO.

To overcome the sparse reward challenge in reinforcement learning (RL) for agents based on large language models (LLMs), we propose Mutual Information Self-Evaluation (MISE), an RL paradigm that utilizes hindsight generative self-evaluation as dense reward signals while simultaneously calibrating them against the environmental feedbacks. Empirically, MISE enables an agent to learn autonomously from dense internal rewards supplementing sparse extrinsic signals. Theoretically, our work provides the first formal foundation for the paradigm of generative self-rewarding. We prove that utilizing hindsight self-evaluation rewards is equivalent to minimizing an objective that combines mutual information with a KL divergence term between the policy and a proxy reward policy. This theoretical insight then informs and justifies our calibration step, which actively aligns these rewards with the optimal policy. Extensive experiments show that MISE outperforms strong baselines, enabling open-source LLMs about 7B parameters to achieve performance comparable to GPT-4o on validation without expert supervision.

Recent advances in Reinforcement Learning with Verifiable Rewards (RLVR) have demonstrated significant potential in single-turn reasoning tasks. With the paradigm shift toward self-evolving agentic learning, models are increasingly expected to learn from trajectories by synthesizing tools or accumulating explicit experiences. However, prevailing methods typically rely on large-scale LLMs or multi-agent frameworks, which hinder their deployment in resource-constrained environments. The inherent sparsity of outcome-based rewards also poses a substantial challenge, as agents typically receive feedback only upon completion of tasks. To address these limitations, we introduce a Tool-Memory based self-evolving agentic framework SEARL. Unlike approaches that directly utilize interaction experiences, our method constructs a structured experience memory that integrates planning with execution. This provides a novel state abstraction that facilitates generalization across analogous contexts, such as tool reuse. Consequently, agents extract explicit knowledge from historical data while leveraging inter-trajectory correlations to densify reward signals. We evaluate our framework on knowledge reasoning and mathematics tasks, demonstrating its effectiveness in achieving more practical and efficient learning.

Decision-making in military aviation Prognostics and Health Management (PHM) faces significant challenges due to the "curse of dimensionality" in large-scale fleet operations, combined with sparse feedback and stochastic mission profiles. To address these issues, this paper proposes Smart Commander, a novel Hierarchical Reinforcement Learning (HRL) framework designed to optimize sequential maintenance and logistics decisions. The framework decomposes the complex control problem into a two-tier hierarchy: a strategic General Commander manages fleet-level availability and cost objectives, while tactical Operation Commanders execute specific actions for sortie generation, maintenance scheduling, and resource allocation. The proposed approach is validated within a custom-built, high-fidelity discrete-event simulation environment that captures the dynamics of aircraft configuration and support logistics.By integrating layered reward shaping with planning-enhanced neural networks, the method effectively addresses the difficulty of sparse and delayed rewards. Empirical evaluations demonstrate that Smart Commander significantly outperforms conventional monolithic Deep Reinforcement Learning (DRL) and rule-based baselines. Notably, it achieves a substantial reduction in training time while demonstrating superior scalability and robustness in failure-prone environments. These results highlight the potential of HRL as a reliable paradigm for next-generation intelligent fleet management.

Reinforcement learning for Large Language Model agents is often hindered by sparse rewards in multi-step reasoning tasks. Existing approaches like Group Relative Policy Optimization treat sampled trajectories as independent chains, assigning uniform credit to all steps in each chain and ignoring the existence of critical steps that may disproportionally impact reasoning outcome. In this paper, we propose T-STAR(Tree-structured Self-Taught Agent Rectification), a framework that recovers the latent correlated reward structure across seemingly independent trajectories. Specifically, we consolidate trajectories into a unified Cognitive Tree by identifying and merging functionally similar steps/nodes. It enables an Introspective Valuation mechanism that back-propagates trajectory-level rewards through the tree to obtain a new notion of variance-reduced relative advantage at step-level. Using the Cognitive Tree, we also develop In-Context Thought Grafting to synthesize corrective reasoning by contrasting successful and failed branches at critical divergence points/steps. Our proposed Surgical Policy Optimization then capitalizes on the rich policy gradient information concentrated at these critical points/steps through a Bradley-Terry type of surgical loss. Extensive experiments across embodied, interactive, reasoning, and planning benchmarks demonstrate that T-STAR achieves consistent improvements over strong baselines, with gains most pronounced on tasks requiring extended reasoning chains.

Reinforcement learning for multi-step reasoning with large language models (LLMs) typically relies on sparse terminal rewards, which creates a poorly conditioned credit-assignment problem: the final feedback is propagated uniformly across all intermediate decisions. This leads to high gradient variance, unstable training, and many ineffective updates, ultimately limiting sustained model improvement. We propose a counterfactual-comparison framework for credit assignment. For each input, the framework samples multiple reasoning trajectories and treats their differences as implicit approximations to alternative decisions. This yields an implicit process-level advantage estimator that converts sparse terminal rewards into step-sensitive learning signals. Building on this framework, we introduce Implicit Behavior Policy Optimization (IBPO), which substantially improves training stability and the performance ceiling on mathematical and code-reasoning benchmarks. Our results point to a promising direction for unlocking the reasoning potential of LLMs.

Medical visual grounding serves as a crucial foundation for fine-grained multimodal reasoning and interpretable clinical decision support. Despite recent advances in reinforcement learning (RL) for grounding tasks, existing approaches such as Group Relative Policy Optimization~(GRPO) suffer from severe reward sparsity when directly applied to medical images, primarily due to the inherent difficulty of localizing small or ambiguous regions of interest, which is further exacerbated by the rigid and suboptimal nature of fixed IoU-based reward schemes in RL. This leads to vanishing policy gradients and stagnated optimization, particularly during early training. To address this challenge, we propose MedLoc-R1, a performance-aware reward scheduling framework that progressively tightens the reward criterion in accordance with model readiness. MedLoc-R1 introduces a sliding-window performance tracker and a multi-condition update rule that automatically adjust the reward schedule from dense, easily obtainable signals to stricter, fine-grained localization requirements, while preserving the favorable properties of GRPO without introducing auxiliary networks or additional gradient paths. Experiments on three medical visual grounding benchmarks demonstrate that MedLoc-R1 consistently improves both localization accuracy and training stability over GRPO-based baselines. Our framework offers a general, lightweight, and effective solution for RL-based grounding in high-stakes medical applications. Code & checkpoints are available at \hyperlink{}{https://github.com/MembrAI/MedLoc-R1}.

Reinforcement Learning (RL) agents often struggle with inefficient exploration, particularly in environments with sparse rewards. Traditional exploration strategies can lead to slow learning and suboptimal performance because agents fail to systematically build on previously successful experiences, thereby reducing sample efficiency. To tackle this issue, we propose a self-imitating on-policy algorithm that enhances exploration and sample efficiency by leveraging past high-reward state-action pairs to guide policy updates. Our method incorporates self-imitation by using optimal transport distance in dense reward environments to prioritize state visitation distributions that match the most rewarding trajectory. In sparse-reward environments, we uniformly replay successful self-encountered trajectories to facilitate structured exploration. Experimental results across diverse environments demonstrate substantial improvements in learning efficiency, including MuJoCo for dense rewards and the partially observable 3D Animal-AI Olympics and multi-goal PointMaze for sparse rewards. Our approach achieves faster convergence and significantly higher success rates compared to state-of-the-art self-imitating RL baselines. These findings underscore the potential of self-imitation as a robust strategy for enhancing exploration in RL, with applicability to more complex tasks.

The process of discovery requires active exploration -- the act of collecting new and informative data. However, efficient autonomous exploration remains a major unsolved problem. The dominant paradigm addresses this challenge by using Reinforcement Learning (RL) to train agents with intrinsic motivation, maximizing a composite objective of extrinsic and intrinsic rewards. We suggest that this approach incurs unnecessary overhead: while policy optimization is necessary for precise task execution, employing such machinery solely to expand state coverage may be inefficient. In this paper, we propose a new approach that explicitly decouples exploration from policy optimization and bypasses RL entirely during the exploration phase. Our method uses a tree-search strategy inspired by the Go-With-The-Winner algorithm, paired with a measure of uncertainty to systematically drive exploration. By removing the overhead of policy optimization, our approach explores an order of magnitude more efficiently than standard intrinsic motivation baselines on hard exploration benchmarks. Further, we demonstrate that the trajectories discovered during exploration can be distilled into deployable policies using existing supervised backward learning algorithms, achieving state-of-the-art performance by a wide margin on Montezuma's Revenge, Pitfall!, and Venture without relying on domain-specific knowledge. Finally, we demonstrate the generality of our framework in high-dimensional continuous action spaces by solving the MuJoCo Adroit dexterous manipulation and AntMaze tasks in a sparse-reward setting, directly from image observations and without expert demonstrations or offline datasets. To the best of our knowledge, this has not been achieved before for the Adroit tasks.

Reinforcement Learning with Verifiable Rewards (RLVR) enhances Large Language Model (LLM) reasoning but is suffer from advantage collapse: when all rollouts of a query receive identical rewards, the group variance vanishes, most damagingly on hard samples, where scaling rollout budgets yields little. We introduce Joint Policy and Prompt Optimization (P O) to mitigate this collapse by alternating continuous policy updates with discrete prompt evolution. P O mines hard samples with a success-rate threshold, evolves reasoning prompts for them with GEPA, and internalizes the elicited trajectories via context distillation, which optimizes each trajectory under the original query and thus removes inference-time prompting, with a Context Ratio Mask (CRM) filtering out extreme likelihood ratios. P O restores critical advantage signals and surpasses the GRPO baseline by up to 8.2 points in average accuracy on six held-out benchmarks across all training datasets and backbones, while also outperforming DAPO and other baselines. The gains are especially pronounced on hard benchmarks, reaching up to 16.3 points above GRPO on average across AIME24 and AIME25. Our findings expose the limits of standard exploration in sparse-reward environments, illuminating the potential of unifying evolutionary algorithms with reinforcement learning. This integration of discrete semantic search and continuous parameter updates provides a self-reinforcing framework that facilitates more effective LLM alignment.

While large language models excel in diverse domains, their performance on complex longhorizon agentic decision-making tasks remains limited. Most existing methods concentrate on designing effective reward models (RMs) to advance performance via multi-turn reinforcement learning. However, they suffer from delayed propagation in sparse outcome rewards and unreliable credit assignment with potentially overly fine-grained and unfocused turnlevel process rewards. In this paper, we propose (HISR) exploiting Hindsight Information to modulate Segmental process Rewards, which closely aligns rewards with sub-goals and underscores significant segments to enhance the reliability of credit assignment. Specifically, a segment-level process RM is presented to assign rewards for each sub-goal in the task, avoiding excessively granular allocation to turns. To emphasize significant segments in the trajectory, a hindsight model is devised to reflect the preference of performing a certain action after knowing the trajectory outcome. With this characteristic, we design the ratios of sequence likelihoods between hindsight and policy model to measure action importance. The ratios are subsequently employed to aggregate segment importance scores, which in turn modulate segmental process rewards, enhancing credit assignment reliability. Extensive experimental results on three publicly benchmarks demonstrate the validity of our method.

Agentic Reinforcement Learning (RL) shows promise for complex tasks, but Text-to-SQL remains mostly restricted to single-turn paradigms. A primary bottleneck is the credit assignment problem. In traditional paradigms, rewards are determined solely by the final-turn feedback, which ignores the intermediate process and leads to ambiguous credit evaluation. To address this, we propose Agentic SQL, a framework featuring a universal two-tiered reward mechanism designed to provide effective trajectory-level evaluation and dense step-level signals. First, we introduce Aggregated Trajectory Reward (ATR) to resolve multi-turn credit assignment. Using an asymmetric transition matrix, ATR aggregates process-oriented scores to incentivize continuous improvement. Leveraging Lyapunov stability theory, we prove ATR acts as an energy dissipation operator, guaranteeing a cycle-free policy and monotonic convergence. Second, Column-Set Matching Reward (CSMR) provides immediate step-level rewards to mitigate sparsity. By executing queries at each turn, CSMR converts binary (0/1) feedback into dense [0, 1] signals based on partial correctness. Evaluations on BIRD show a 5% gain over binary-reward GRPO. Notably, our approach outperforms SOTA Arctic-Text2SQL-R1-7B on BIRD and Spider 2.0 using identical models, propelling Text-to-SQL toward a robust multi-turn agent paradigm.

Reinforcement Learning with Verifiable Rewards improves reasoning in large language models, yet on-policy learning often suffers from cold-start challenges in sparse-reward settings. Recent mixed-policy approaches address this by combining off-policy teacher data with on-policy training. However, simply combining these introduce a persistent off-policy gradient mass that risks training collapse and instability. To address this challenge, we propose Hindsight-Anchored Policy Optimization (HAPO), a framework that allows teacher intervention to act as a temporary support. HAPO employs Beta-Binomial confidence gating, an adaptive gating mechanism that decides when to open the gate for teacher intervention. The intervention operates with Synthetic Success Injection, which replaces the group's lowest-reward rollout with a verified teacher trajectory. We also introduce adaptive threshold annealing, which gradually retracts the support and restores on-policy training within a finite horizon to mitigate persistent off-policy drift. We demonstrate that HAPO can be layered on top of existing mixed-policy methods in a generalizable manner. Across six math reasoning benchmarks and two model scales, HAPO improves the average accuracy of three major mixed-policy methods while maintaining training stability.

Search-augmented large language models (LLMs) trained with reinforcement learning (RL) have achieved strong results on open-domain question answering (QA), but training still remains a significant challenge. The optimization is often unstable due to sparse rewards and difficult credit assignments across reasoning and tool calls. To address this, we introduce Turn-Level Information Potential Reward Shaping (TIPS), a simple framework that assigns dense, turn-level rewards to each reasoning + tool-call segment based on the increased likelihood of the correct answer under a teacher model. By leveraging the potential-based reward shaping, TIPS offers fine-grained and policy-invariant guidance that overcomes the limitations of outcome-only optimization. Evaluated on seven QA benchmarks, TIPS consistently outperforms GRPO/PPO baselines and substantially improves training stability. For instance, with a Qwen-2.5 7B Instruct model, TIPS improves the average Exact Match score by 11.8% and F1 by 13.6% relative to PPO. Our results demonstrate that turn-level information-potential reward shaping provides an effective and general solution to sparse-reward credit assignment for multi-turn LLM reasoning.

In Reinforcement Learning with Verifiable Rewards (RLVR), constructing a robust advantage baseline is critical for policy gradients, effectively guiding the policy model to reinforce desired behaviors. Recent research has introduced Generalist Value Models (such as $V_0$), which achieve pre-trained value estimation by explicitly encoding model capabilities in-context, eliminating the need to synchronously update the value model alongside the policy model. In this paper, we propose $V_{0.5}$, which adaptively fuses the baseline predicted by such value model (acting as a prior) with the empirical mean derived from sparse rollouts. This constructs a robust baseline that balances computational efficiency with extremely low variance. Specifically, we introduce a real-time statistical testing and dynamic budget allocation. This balances the high variance caused by sparse sampling against the systematic bias (or hallucinations) inherent in the value model's prior. By constructing a hypothesis test to evaluate the prior's reliability in real-time, the system dynamically allocates additional rollout budget on demand. This mechanism minimizes the baseline estimator's Mean Squared Error (MSE), guaranteeing stable policy gradients, even under extreme sparsity with a group size of 4. Extensive evaluations across six mathematical reasoning benchmarks demonstrate that $V_{0.5}$ significantly outperforms GRPO and DAPO, achieving faster convergence and over some 10% performance improvement.

We introduce Reward-Zero, a general-purpose implicit reward mechanism that transforms natural-language task descriptions into dense, semantically grounded progress signals for reinforcement learning (RL). Reward-Zero serves as a simple yet sophisticated universal reward function that leverages language embeddings for efficient RL training. By comparing the embedding of a task specification with embeddings derived from an agent's interaction experience, Reward-Zero produces a continuous, semantically aligned sense-of-completion signal. This reward supplements sparse or delayed environmental feedback without requiring task-specific engineering. When integrated into standard RL frameworks, it accelerates exploration, stabilizes training, and enhances generalization across diverse tasks. Empirically, agents trained with Reward-Zero converge faster and achieve higher final success rates than conventional methods such as PPO with common reward-shaping baselines, successfully solving tasks that hand-designed rewards could not in some complex tasks. In addition, we develop a mini benchmark for the evaluation of completion sense during task execution via language embeddings. These results highlight the promise of language-driven implicit reward functions as a practical path toward more sample-efficient, generalizable, and scalable RL for embodied agents. Code will be released after peer review.

Large language models (LLMs) typically receive diverse natural language (NL) feedback through interaction with the environment. However, current reinforcement learning (RL) algorithms rely solely on scalar rewards, leaving the rich information in NL feedback underutilized and leading to inefficient exploration. In this work, we propose GOLF, an RL framework that explicitly exploits group-level language feedback to guide targeted exploration through actionable refinements. GOLF aggregates two complementary feedback sources: (i) external critiques that pinpoint errors or propose targeted fixes, and (ii) intra-group attempts that supply alternative partial ideas and diverse failure patterns. These group-level feedbacks are aggregated to produce high-quality refinements, which are adaptively injected into training as off-policy scaffolds to provide targeted guidance in sparse-reward regions. Meanwhile, GOLF jointly optimizes generation and refinement within a unified RL loop, creating a virtuous cycle that continuously improves both capabilities. Experiments on both verifiable and non-verifiable benchmarks show that GOLF achieves superior performance and exploration efficiency, achieving 2.2$\times$ improvements in sample efficiency compared to RL methods trained solely on scalar rewards. Code is available at https://github.com/LuckyyySTA/GOLF.

Reward Machines (RMs) are an established mechanism in Reinforcement Learning (RL) to represent and learn sparse, temporally extended tasks with non-Markovian rewards. RMs rely on high-level information in the form of labels that are emitted by the environment alongside the observation. However, this concept requires manual user input for each environment and task. The user has to create a suitable labeling function that computes the labels. These limitations lead to poor applicability in widely adopted RL frameworks. We propose Symbolic Reward Machines (SRMs) together with the learning algorithms QSRM and LSRM to overcome the limitations of RMs. SRMs consume only the standard output of the environment and process the observation directly through guards that are represented by symbolic formulas. In our evaluation, our SRM methods outperform the baseline RL approaches and generate the same results as the existing RM methods. At the same time, our methods adhere to the widely used environment definition and provide interpretable representations of the task to the user.

Reinforcement learning (RL) for mathematical reasoning can suffer from reward sparsity: for challenging problems, LLM fails to sample any correct trajectories, preventing RL from receiving meaningful positive feedback. At the same time, there often exist human-written reference solutions along with the problem (e.g., problems from AoPS), but directly fine-tuning on these solutions offers no benefit because models often cannot imitate human proofs that lie outside their own reasoning distribution. We introduce Reference-Guided Fine-Tuning (ReGFT), a simple and effective method that utilizes human-written reference solutions to synthesize positive trajectories on hard problems and train on them before RL. For each problem, we provide the model with a partial reference solution and let it generate its own reasoning trace, ensuring the resulting trajectories remain in the model's reasoning space while still benefiting from reference guidance. Fine-tuning on these reference-guided trajectories increases the number of solvable problems and produces a checkpoint that receives more positive rewards during RL. Across three benchmarks (AIME24, AIME25, BeyondAIME), ReGFT consistently improves supervised accuracy, accelerates DAPO training, and raises the final performance plateau of RL. Our results show that ReGFT effectively overcomes reward sparsity and unlocks stronger RL-based mathematical reasoning.