Reinforcement learning from verifiable rewards (RLVR) drives chain-of-thought reasoning in large language models, yet its binary outcome reward cannot distinguish among correct trajectories. Existing dense reward alternatives, from surface heuristics to process reward models, either ignore the expert solutions already present in training corpora or require expensive offline annotation. We propose Gradient-Aligned Reward (GAR), which operates in the policy's own gradient space: truncated backpropagation through the output projection layer extracts a compact gradient vector for each rollout, and cosine similarity with an expert-anchor gradient yields a dense, reasoning-aware reward with less than 9% wall-clock overhead. We prove that this cosine admits a multiplicative decomposition into prediction-error and activation-pattern factors, providing a concrete characterization of what the alignment signal measures. On Qwen3-4B and Qwen3-8B, GAR consistently improves over GRPO and other baselines on competition-level math benchmarks and transfers to GPQA Diamond and MMLU-Pro without domain-specific data. Code and data are available at https://github.com/LQgdwind/GAR.
Reinforcement learning with verifiable rewards (RLVR) has emerged as a powerful paradigm for large language model (LLM) post-training, but its reliance on coarse outcome rewards leads to limited guidance on intermediate reasoning processes. Existing approaches such as process reward modeling and on-policy distillation introduce additional constraints, such as reliance on a specialized reward model or assuming identical reasoning patterns between teacher and student. Nevertheless, we observe that once a reasoning process first goes wrong, evaluating the subsequent reasoning provides limited additional information, as it is already conditioned on an invalid prefix. Therefore, we propose Cliff, a reward shaping strategy that utilizes an off-the-shelf LLM as a teacher to identify the first mistake in each rollout. As a result, the rollout is naturally decomposed into two parts: a correct prefix and an incorrect suffix. Cliff then converts this signal into token-level advantages, assigning positive advantages for the correct prefix and negative feedback afterward. Experiments across 12 different scenarios demonstrate that Cliff consistently improves reasoning performance, outperforming on-policy distillation by 15% and standard GRPO by 7%, even with teachers of modest capability. Furthermore, we analyse the role of ``ground truth'' in Cliff and investigate its training dynamics. These results establish Cliff as a simple, general and effective approach for improving RLVR with richer, fine-grained supervision.
Chenyu Zhou, Qiliang Jiang, Shuning Wu +1cs.LG cs.AI
Multi-turn agentic RL increasingly treats credit assignment as a targeting problem: given a terminal verifiable reward, per-turn methods localize credit onto the turns that mattered. We identify the structural quantity that predicts when this is the right move, the verifier information density V_d = k/C (the fraction of an agent's C-step causal chain whose per-turn correctness the verifier exposes), and show that terminal-state verifiers sit deep in a low-V_d regime where targeting is the wrong axis. In controlled shared-rollout comparisons on tau^2-bench that separate reward density from credit geometry, a continuous dense reward spread uniformly beats the sparse binary outcome reward (net-harmful on 4/5 seeds), while concentrating the same advantage on progress turns or on random turns is equally harmful: targeting is second-order. The mechanism is coverage: terminal-state verification collapses the observable signal to a single final-write turn (k=1 in 98% of rollouts) while success requires a 5-8 step chain of prerequisite tool calls. A synthetic phase boundary places the crossover at V_d* ~ 0.8, whereas measured V_d is ~0.15 on tau^2-bench and ~0.4 on BFCL V3; uniform also wins on BFCL, where a matched-concentration shuffled control is negative on 8/8 seeds. The effect reproduces across model families on ToolACE-2-8B (Delta = -0.048 over 32 pre-registered seeds; an independent 20-seed replication is itself significant), and a pre-registered matched-budget breadth sweep traces a monotone dose-response whose deficit vanishes only at full chain coverage, with a reward-to-go arm reaching full-coverage parity. Uniform redistribution is the zero-information coverage default that per-turn schemes must beat; we contribute the matched-concentration shuffled control that any targeting claim should clear.
Goal-conditioned reinforcement learning struggles with long horizons when rewards are sparse. While a planner can provide subgoals to guide a low-level policy, its use at test time may introduce practical subgoal management difficulties. An alternative paradigm utilizes a high-level planner to assist learning, while the policy remains conditioned only on the final goal, enabling planner-free deployment. Among these methods, Reinforcement Learning with Imagined Subgoals (RIS) introduces a regularization term that encourages the policy to take the same actions for the final goal as it does for an intermediate goal. This regularization, however, may lead to goal-chaining issues when intermediate goals are low-dimensional. Potential-based reward shaping (PBRS) translates plans into an additional reward while ensuring that the optimal policy remains unchanged. Yet, it can generate deceptive rewards in terminal states. We study these failure cases and first propose an alternative reward shaping method (RS) that removes these deceptive rewards at the expense of theoretical guarantees of PBRS. Similar to this RS variant, we then propose another method named Locally-Guided Actor Critic (LG-AC) that rewards the agent for reaching intermediate goals. Unlike RS, where intermediate rewards are implicit in the shaping signal, we explicitly condition a value estimator on the full sequence of intermediate goals but represent the value function as a sum of subgoal-conditioned value functions, enabling dense hindsight relabeling. We evaluate all these methods in tasks with challenging goal-chaining requirements and empirically highlight specific cases in which either action regularization or reward shaping yield low performance, while LG-AC achieves the best overall performance across tasks.
Christophe D. Hounwanou, John Emeka Eze, Yaé U. Gabacs.LG cs.AI
Combining large language models with reinforcement learning is increasingly explored, yet the theoretical status of LLM-derived reward signals is often left implicit. We formalize the hybrid LLM-planner and RL-controller architecture as a Goal-Augmented Markov Decision Process and show that when the LLM per-state progress score is used as a bounded potential function, the resulting shaping term preserves the optimal policy set even when the LLM scores are inaccurate. This guarantee is stronger than what general LLM-as-reward approaches provide. We verify the result numerically on a small MDP under four potential configurations, including an adversarial one scaled to twenty times the base reward magnitude.
Signal temporal logic (STL) provides a formal language for specifying real-time properties of real-valued observations, along with a quantitative robustness score for monitoring satisfaction. Control synthesis from STL specifications is of interest since manual controller design becomes infeasible as real-world systems grow in complexity. Moreover, many modern autonomous and AI-enabled systems lack accurate and complete system models, which makes optimization-based synthesis approaches unsuitable and motivates learning-based control. Prior work uses STL robustness scores as rewards in reinforcement learning (RL) to obtain control policies satisfying given specifications; however, robustness depends on execution history, leading to intractable state space expansion for general long-horizon specifications with arbitrarily nested temporal operators. This work introduces a novel automata-based approach that provides an efficient memory mechanism and associated Markovian rewards suitable for RL frameworks. Our approach constructs a timed alternating automaton from the given STL specifications, augments the state space with automaton locations and clock valuations, and derives rewards from the automaton acceptance condition. We empirically demonstrate that our approach learns policies that achieve higher robustness scores and satisfaction rates than those learned by existing approaches using robustness-based rewards.
Instant delivery platforms have become a critical component of urban logistics, increasingly relying on crowdsourced couriers to fulfill highly dynamic orders. In real-world systems, couriers are not exclusive to a single platform and may concurrently serve multiple platforms, while each platform can only observe its own orders and couriers' interactions due to privacy and operational constraints. This results in a multi-platform dispatch environment with inherent partial observability. However, most existing works on dispatch optimization assume full courier observability and mandatory assignment acceptance, causing substantial performance degradation when deployed in realistic multi-platform settings. In this paper, we propose POLO, a partially observable multi-agent reinforcement learning framework for dispatching optimization in multi-platform instant delivery systems. POLO firstly models each platform-grid pair as an independent agent that learns dispatch policies solely from platform-local observations, aligning the learning process with real-world privacy and operational constraints. To support effective decision-making under incomplete and heterogeneous courier information, POLO introduces a novel attention-based policy representation that selectively aggregates inter-courier information. Moreover, we design a counterfactual reward shaping mechanism to mitigate the non-stationarity induced by joint actions across grids, leading to more stable and scalable learning. We develop a high-fidelity simulator to evaluate dispatch performance under varying numbers of platforms and system scales. Extensive experiments demonstrate that POLO consistently outperforms strong baselines in terms of platform revenue and courier travel efficiency, highlighting its robustness and effectiveness in realistic multi-platform settings.
Sparse, delayed, and weakly informative rewards remain central obstacles to efficient reinforcement learning. Reward shaping addresses these limitations by supplementing the task reward with an auxiliary signal that can accelerate learning while, in the classical setting, the original objective remains the evaluation criterion. Established theory guarantees safety for fixed shaping signals: potential-based reward shaping preserves optimal policies when the auxiliary term is the discounted difference of a time-invariant potential. In contemporary reinforcement learning systems, however, both the learner and the information available for guidance evolve during training: value estimates improve, novelty diminishes, feedback shifts, and predictive models are refined. Adaptive reward mechanisms occur across exploration, Bayesian inference, human-in-the-loop learning, automated reward design, and foundation-model-based approaches. This study introduces a unified analytical framework for comparing dynamic reward shaping and neighbouring adaptive reward mechanisms. The proposed framework distinguishes parametric revision from state-dependent variation, separates additive shaping from reward replacement and reward-adjacent guidance, and organises existing methods along temporal, informational, and theoretical dimensions. Using this framework, twelve method families are comparatively analysed. The framework further highlights the conditions under which optimality guarantees survive contemporary deep reinforcement learning pipelines, replay buffers, bootstrapped critics, and reward normalisation, while exposing the unresolved relationship between adaptation rate and learner stability.
Cooperative multi-agent reinforcement learning often adds social terms to individual rewards, yet the scale of those terms is usually chosen by hand. We ask whether a guilt signal can instead be calibrated from human neural and behavioural data and transferred to artificial agents. Using the public SoDec responsibility fMRI dataset (40 participants), we fit a subject-fixed-effects regression of momentary-happiness changes on outcome-type counts and recover a guilt weight as the Partner-negative minus Social-negative contrast ($\hat{w}=1.118$, Cohen's $d=0.214$). We embed this weight in a two-agent Social Lottery environment and train independent Proximal Policy Optimization actor-critics under four shaping regimes: neurally calibrated, uniform constant, zero (selfish), and a unit-coefficient oracle. Across 1{,}000 evaluation episodes per condition, the calibrated agents track the human Social safe-choice rate most closely ($0.459$ vs.\ human $0.484$; $\mathrm{KL}=0.0012$), while the other three conditions deviate by one to three orders of magnitude in KL. Human neurobehavioural priors can therefore act as quantitative constraints on prosocial reward shaping.
Agentic reinforcement learning enables LLM agents to learn through interaction, but sparse trajectory-level rewards reveal success without identifying which intermediate decisions deserve credit. Training-only privileged skills can provide denser supervision by allowing the same frozen policy snapshot to rescore fixed tokens from skill-free trajectories while conditioned on task-matched procedural skills. Existing methods, however, do not jointly calibrate teacher scores across interaction steps, relate teacher confidence to realized returns, and integrate the resulting signal into native reward-to-advantage construction. We introduce Agentic Reinforcement Learning with Self-Distilled Reward Shaping (ADRS), a framework for constructing return-associated token-level credit for multi-turn language agents. ADRS centers and normalizes privileged token scores within each step, modulates them with a return-associated Teacher Value Advantage (TVA) gate based on within-group confidence--return association, and incorporates the gated token signal into native RL credit construction. Together, these components determine what the teacher prefers, when that preference is return-relevant, and how it enters the native reinforcement-learning credit path, while keeping rollouts and inference skill-free. Finally, experiments across three interactive benchmarks show that ADRS consistently improves performance on long-horizon tasks, with gains persisting across RL backbones, reduced-data settings, unseen tasks, and extended training. For anonymous review, our code is available at the following the link: https://github.com/gitrxh/ADRS-arxiv
Credit assignment is a fundamental challenge in cooperative multi-agent reinforcement learning, particularly in embodied AI settings characterized by limited and delayed feedback as well as dynamically changing numbers of active agents. We propose MARS-RA, a framework that reformulates credit assignment as a rank aggregation problem using contribution-based pairwise comparisons among agents generated by large multimodal models. This shift from absolute to relative estimation ensures robustness against noise and dynamic agent participation, converting comparison results into contribution scores for potential-based reward shaping. We provide theoretical justification for the convergence and robustness of the proposed framework, and show that Shapley values can be used as an interpretive reference. Experimental results on challenging tasks of different types indicate that MARS-RA can guide agents toward effective cooperation.
Reinforcement Learning from Human Feedback (RLHF) is the standard approach for aligning large language models with human preferences, but its quality is limited by static, task-agnostic reward models. This mismatch leads to sparse learning signals and suboptimal alignment. We introduce MeRLa (Meta-Learned Reward Shaping), a principled framework that meta-learns a task-aware shaping function $Φ(x,y;φ)$ across auxiliary tasks before RLHF training. The learned shaping produces a composite reward that preserves policy optimality while providing task-specific learning signals. Our meta-objective combines task discrimination, entropy regularization, and potential-based conservation for stable convergence. We provide theoretical guarantees for policy invariance, analyze representation drift sensitivity, and formally address incentive misalignment from entropy maximization. Experiments on LLaMA-3-8B across four benchmarks show consistent improvements over PPO, DPO, GRPO, and DAPO, achieving a 90.8% length-controlled win rate on AlpacaEval 2.0 and a score of 9.14 on MT-Bench, with 41% less training instability. MeRLa retains its benefits when combined with process-based and rubric-based enhanced rewards.
Designing effective reward functions for model-free reinforcement learning under non-holonomic constraints remains a persistent challenge, often resulting in severe local minima such as policy paralysis or over-conservative hazard avoidance. In this work, we present a parameterized reward shaping framework featuring coverage-gated alignment feedback, drive-direction switch regularization, and an aligned episode termination mechanism evaluated on an autonomous parallel parking task. Crucially, we show that environmental reward parameters and algorithmic hyperparameters are deeply co-dependent, requiring joint meta-optimization to achieve stable convergence. By employing surrogate-based Bayesian optimization, our co-optimized Deep Q-Network (DQN) agent resolves characteristic control failure modes, significantly outperforming uncalibrated baselines across both success rate and trajectory smoothness.
Dense per-step supervision is an appealing remedy for sparse-reward, long-horizon LLM agents: reward the agent for predicting its next observation, and memory should follow. We show that under group-normalized RL (GRPO), this recipe does not merely fail -- it destroys the policy. Across Qwen3-1.7B/4B/8B on ALFWorld, a potential-based prediction reward drives every run into a degenerate absorbing state (prediction accuracy -> 1.0, task success -> 0,episode length pinned at the horizon): the "dark room" pathology, built automatically by the optimizer. A single-factor ablation localizes the cause -- removing only GRPO's std normalization turns the same reward from catastrophic (0%) into baseline parity -- and a two-line proposition explains why: in all-fail groups the z-scored advantage is invariant to the shaping coefficient, so bounded rewards become unbounded pressure and annealing cannot help. Our central insight generalizes this: what z-scoring amplifies is a dense signal's within-group variance while all-fail groups dominate, so signals whose variance decays by mastery are structurally amplifier-safe.This variance-profile criterion retrodicts our collapses, carries preregistered predictions for arms that had not yet run, and is consistent with published reward-channel successes (a compatibility check, not an independent test). Finally, a controlled signal-delivery matrix (identical signal, varying only the consumption mechanism) shows the reward channel is at best neutral while the auxiliary-loss channel gains ~20 points -- and a shuffled-gold placebo matches the true-gold arm, so the gap survives without correct labels. Endpoints are single-seed; seed replication and group-size controls are preregistered and in progress.
Knowledge graphs (KGs) are widely used to inject prior knowledge into reinforcement learning (RL), yet the literature is dominated by single-domain, positive-result method papers, so we lack a systematic account of when KG structure helps an agent, when it is neutral, and when it hurts. We conduct a controlled study that independently varies the RL task, the injection mechanism (state features, action masking, or potential-based reward shaping), and KG quality. Using a synthetic, fully controllable KG over MiniGrid environments, we report three findings. First, on compositional sparse-reward tasks structured KG guidance improves sample efficiency and solve reliability (70% to 97% of seeds), and a shuffle control that permutes the KG's edges while preserving their count collapses the benefit toward baseline (masking p=0.0001; shaping p=0.006), so the gain is structural rather than generic regularization. Second, KG value scales with the amount of task-relevant knowledge the graph contains. Third, and most consequential, safety depends on the mechanism: soft, optimality-preserving injection benefits from correct knowledge and harmlessly ignores incorrect knowledge, whereas hard masking is brittle, forbidding essential actions when the KG is incomplete or corrupted and making a wrong KG worse than none. A UMLS-derived clinical case study on sepsis management under offline RL is a careful null, underscoring that benefits require task structure the chosen mechanism can exploit. Our results give practitioners concrete guidance on how, and how much, to trust a KG when using it to guide RL.
Reinforcement learning for deep-search agents has largely focused on trajectory-level scoring -- outcome correctness, citation-aware rewards, and evidence coverage. Yet the actions that expose supporting documents receive no targeted credit, a gap we call the reward-credit mismatch. We propose STAMP, in which a reference-based verifier judges whether each cited document supports an entity or relation in a training-time evidence graph, and first-exposure attribution traces each supported citation back to the action that first surfaced it. This step credit is injected through sign-preserving advantage modulation, which redistributes advantage across steps without changing the trajectory-level reward or the relative ranking of trajectories within each group. On BrowseComp, BrowseComp-ZH, and xbench-DS, STAMP improves the GRPO baseline by +2.0/+5.5/+3.0 points under matched SFT initialization, training data, and search tools, and composes with both outcome-only and citation-rubric base rewards. Component ablations confirm that the provenance-based credit signal and the sign-preserving advantage modulation each contribute to the gains.
Inducing cooperation among distributed agents is still a difficult problem in the field of multi-agent reinforcement learning (MARL), particularly in social dilemma situations. There, individual interests are misaligned with the common good and individual rationality leads to suboptimal group outcomes. In contrast, humans are able to achieve cooperation with one another in such situations. A common explanation for such cooperative behavior is that individuals have social preferences. In order to achieve cooperation in MARL, we design a new utility function integrating altruistic preferences (incentive for other's reward) and fairness preferences (incentive for equality) from social psychology and behavioral economics, namely, Altruistic and Fairness Preference (AFP), a reward-sharing mechanism which converts one's own and other's rewards to incentives for cooperative behavior. We performed comparative experiments with standard RL and inequity aversion agents in two challenging sequential social dilemma games, and showed that AFP agents successfully achieved mutual cooperation with more collective rewards and higher equity than the baselines. To further understand the progression of AFP during training, we subsequently explore the effects of altruistic preferences and fairness preferences on agents' behavior. The results suggest that altruistic preferences encourage agents to contribute to the public goods, and fairness preferences induce mutual behavior between agents.
Large Language Models (LLMs) offer a natural interface for translating human objectives into reward signals for cooperative multi-agent reinforcement learning (MARL), yet the training-time dynamics of this integration remain poorly understood. We show that dynamically updating LLM-generated reward weights during off-policy MARL violates the stationarity assumption of Potential-Based Reward Shaping (PBRS) and contaminates the experience replay buffer, whose stored transitions carry reward labels computed under stale shaping weights. We characterise the result as a regime-dependent failure whose severity depends on how competent the unshaped baseline already is. To control it we propose two stabilisation strategies: a Phase-Based Freeze Schedule that enforces strict stationarity within training phases, and Exponential Moving Average (EMA) smoothing that bounds per-episode weight drift. We evaluate across three cooperative environments and five random seeds with QMIX, complemented by an exploratory VDN extension, yielding a three-regime taxonomy. In the augmentative regime (Simple Spread), where the baseline is functional (74.4 %), EMA significantly improves success to 86.7 % ($+12.3$ pp, $p<0.01$) while naive dynamic updates collapse it to 15.2 %. In the essential regime (Level-Based Foraging), where the baseline is broken (0.1 %), any shaping unlocks the task (95.9 % under EMA). In the supplementary regime (SMAC 3m), where the baseline is near-saturated (98.8 %), stabilised shaping preserves performance (99.9 %) while unstabilised shaping adds variance without gain. These findings establish reward-signal stationarity as a necessary design constraint and indicate that regime placement is a practical predictor of whether dynamic LLM shaping helps or harms.
We study how to predict the downstream closed-loop performance of a learned latent world model from validation-time diagnostics alone. Choosing the right checkpoint from a world-model training run is difficult: validation loss and multi-step prediction RMSE keep improving long after closed-loop performance has collapsed. We present a suite of structural validation-time diagnostics drawn from optimal-control theory and apply them to Gymnasium's LunarLander v3, which features shaped rewards. We train an RSSM [5, 4] world model on it and treat per checkpoint CEM-MPC return as the oracle for closed-loop quality. By evaluating 40 metrics against this oracle, we find that the strongest single predictor is the Reward Observability Fraction (ROF), which measures the reward predictor's dependence on the observable subspace. We combine ROF with three structural regularizers into a single-number offline checkpoint selection score, the Composite Reward Observability Fraction (CROF). The CROF-selected world model trains a model-based A2C policy that beats a fairly evaluated model-free A2C baseline by ~24.5 return points while using ~65x fewer real-environment interactions, and the same world model also drives a strong zero-shot CEM-MPC policy. Code and data: https://github.com/nsmoly/LunarLander_RSSM.
Jan Stenner, Alexander Kilian, Sebastian Peitz +1cs.LG
This paper studies Reinforcement Learning as an online controller for curtailment-aware workload shifting in wind-turbine-integrated high-performance computing (HPC) data centers. We introduce a reproducible fixed-day simulation framework with synthetic wind and price signals and delayed completion feedback, designed to be extensible toward more complex scenarios. As a controlled benchmarking basis, we then focus on the minimal case with one wind turbine and one co-located data center. In this setting, pure Reinforcement Learning exhibits a pronounced credit-assignment problem and tends to underuse free wind energy early in the day. We therefore evaluate two complementary countermeasures: optimization-based Imitation Learning and potential-based Reward Shaping. Across multi-seed training and a 200-day test set, Proximal Policy Optimization (PPO) and a Soft Actor-Critic (SAC) variant with an additional on-policy update routine achieve strong empirical performance among learned policies, and both Imitation Learning and Reward Shaping provide improvements in relevant configurations. A performance gap to the optimizer remains, which is expected: the optimizer plans offline with full-day foresight, whereas Reinforcement Learning must decide online from current observations without future realizations. The benchmark and ablation results provide a transparent basis for extending the approach toward richer multi-site and continuous-time scenarios.
Group Relative Policy Optimization (GRPO) is a default recipe for process-supervised reinforcement learning of LLM reasoners, and dense process supervision -- via learned process reward models (PRMs) or on-policy-distillation KL signals -- is a common way to densify its otherwise weak outcome reward. Layering such a step-level signal on top of GRPO's group-standardized advantage, however, exposes three structural pathologies: \emph{channel contamination} between the pooled process, outcome, and format streams at group standardization; \emph{resolution mismatch} between the granularity of the process signal and the granularity of the logical decisions being credited; and a \emph{cumulative trap} by which GRPO's return-to-go sum surfaces either length inflation or truncated exploration depending on the sign regime of the signal. We propose \textbf{PASS} (\emph{Process Advantage Signal Shaping}), a compact middleware that sits between any scalar step-level process signal and GRPO's clipped surrogate and addresses the three pathologies in turn: \emph{Advantage Fusion} standardizes the three streams independently within each group, \emph{Chunk-by-Value} derives value-homogeneous chunks from the signal itself and broadcasts credit within each chunk, and \emph{Divide-Length} converts the cumulative objective into an average-value-density score. We validate PASS across two domains and two process-signal paradigms -- a learned PRM on mathematical reasoning and an on-policy-distillation KL signal (with a generalized variant) on multi-hop question answering -- and under two group-standardization operators. In every regime PASS delivers a consistent pass@1 gain over the corresponding GRPO baseline.
Reinforcement learning with verifiable rewards (RLVR) for training LLMs typically rely on ground-truth answers to assign rewards, limiting their applicability to tasks where the ground-truth solution is unknown. We introduce a \textbf{R}anking-\textbf{i}nduced \textbf{VER}ifiable framework (RiVER) that trains LLMs on score-based optimization tasks without ground-truth solutions, using deterministic execution feedback as continuous-valued supervision. When applying group-relative RL to such continuous rewards, we identify two key challenges: \emph{scale dominance}, where uncalibrated score magnitudes across test instances distort policy updates, and \emph{frequency dominance}, where repeatedly sampled suboptimal solutions can outweigh rare but stronger candidates. RiVER addresses these challenges with calibrated reward shaping that uses instance-wise comparisons and emphasizes top-ranked solvers while retaining bounded feedback for other valid solutions. We train on 12 AtCoder Heuristic Contest tasks and evaluate on Algorithm Engineering Benchmark (ALE-Bench), LiveCodeBench, and USACO. RiVER advances Qwen3-8B and GLM-Z1-9B-0414 by 8.9\% and 9.4\% in ALE rating rank. More importantly, despite training exclusively on score-based tasks without any ground-truth solutions, RiVER also improves the backbones across exact-solution benchmarks such as LiveCodeBench and USACO by an absolute average improvement of 2.4\% and 3.5\%. By contrast, baselines trained with raw execution scores improve ALE rating but fail to transfer to exact-solution benchmarks. These results suggest that score-based optimization tasks, combined with proper reward calibration, can serve as effective training environments for general coding ability without ground-truth solutions.
Job-search platforms rely on low-bandwidth query interfaces that often fail to capture the high-dimensional complexity of candidate profiles. We present an end-to-end RLAIF (Reinforcement Learning from AI Feedback) framework to generate \emph{portable} job search queries, terms that abstract away seeker-specific identifiers while preserving generalizable qualifications. This task introduces a highly adversarial reward surface where policy optimization frequently exploits flaws in LLM-as-judge rubrics, resulting in degenerate verbatim-copying behaviors. We conducted comprehensive empirical experiments to isolate the impact of optimization mechanics against structured reward engineering. Our results demonstrate that for critic-free optimizers, performance is overwhelmingly dictated by robust reward shaping, rendering the specific choice of algorithm largely immaterial. While critic-free per-rollout baseline methods (RLOO and REINFORCE++) natively resist reward-hacking, the group-relative advantage normalization in GRPO appears uniquely sensitive to spurious reward signals, making it disproportionately susceptible to exploitation. We show that introducing a deterministic, rule-based reward floor to correct for rewards assigned to verbatim copying mitigates this failure mode, resulting in a substantial $+0.147$ quality improvement on a cross-family evaluation judge. Ultimately, we show that the training-time reward model inflates performance gains by $2.4\times$, confirming that the training success is fundamentally dependent on enforcing reward-shaping disciplines rather than selecting alternative optimizers.
Sparse rewards are inherently challenging for reinforcement learning agents as they lack intermediate feedback to guide exploration and to correctly attribute the sparse success rewards to relevant parts of the trajectory. Naive reward shaping can induce reward hacking, yielding policies that exploit auxiliary signals instead of solving the intended task. Potential-based reward shaping (PBRS) guarantees preservation of the optimal policy set, but requires the definition of a heuristic potential function over the state space. In this work, we introduce the VLM-guided PBRS framework VLM-PBRS that learns the potential function directly from vision language model (VLM) feedback. We query a lightweight VLM to obtain preferences over image pairs and train a model of the potential function using these preferences. As this approach is based on potential-based reward shaping, it preserves the original optimal policies, and removes the need for expert-designed reward shaping terms. Because large VLMs are prohibitively expensive to invoke repeatedly during policy learning, we employ smaller, more computationally efficient VLMs. Although the resulting preference labels are less accurate, empirical evidence shows that the preference labels can still be used to accelerate learning. We validate our method empirically in the Meta-World and Franka Kitchen environments and highlight the connection between VLM preference label accuracy and sample efficiency improvements. Our contributions are threefold: (1) the first application of VLM preference-based learning to synthesize a potential function for PBRS, (2) a principled, low-cost solution that leverages small VLMs, and (3) extensive empirical demonstration of improved sample efficiency and robustness to reward hacking.
Transfer learning improves policy learning efficiency by reusing knowledge from source tasks, providing a feasible paradigm for safe and efficient autonomous highway lane changing decision-making. Existing methods frequently encounter transfer mismatch induced by distribution shifts between source and target domains, leading to training oscillation and performance decline. Besides, target domain adaptation depends on exploratory interactions, which struggles to guarantee training safety in safety-critical lane changing cases. To tackle these limitations, this paper proposes a safe transfer reinforcement learning framework for autonomous highway lane changing. First, we design an adaptive teacher intervention mechanism based on instantaneous safety cost to restrain risky exploration and fade intervention strength progressively, with theoretical analysis on return bounds for mixed behavior policy. This intervention also produces dual-source samples for joint training. Second, a teacher-guided safe transfer module embeds action evaluation information of teacher policy into student learning via reward shaping to boost training safety and efficiency, with teacher guidance decaying as policy safety rises. Third, a teacher-guided weighted optimization mechanism adjusts sample weights in policy optimization using a likelihood ratio factor to stabilize transfer performance. Experiments under varied traffic densities and validations on real-world NGSIM dataset reveal that our method surpasses baseline approaches by over 52.2% in safety and 5.0% in learning efficiency. Results verify the efficacy and robustness of our safety-aware transfer strategy for autonomous highway lane changing under various traffic conditions.
Group-based reinforcement learning effectively post-trains LLM agents for long-horizon, sparse-reward tasks by deriving step-level credit from trajectory outcomes. However, this ties a step's credit to its rollout's final outcome: semantically near-identical intermediate steps receive opposite credit depending on whether their trajectory eventually succeeded or failed. Such semantic credit inconsistency sends conflicting gradients to similar actions and wastes the partially-correct progress inside failed rollouts. Motivated by this, we propose Semantic Consistency Policy Optimization (SCPO), a value-free reward-shaping method that mitigates this inconsistency by recovering step-level credit from successful siblings in the same rollout group. Concretely, SCPO scores each failed step against a successful sibling and adds positive step-level credit for new progress along that sibling. On ALFWorld and WebShop, SCPO matches or exceeds strong group-based baselines, reaching 93.7+/-4.1 percent success on ALFWorld and 74.8+/-2.0 percent on WebShop at 1.5B parameters, with gains concentrated on the hardest multi-step tasks.
Raymond Tsao, Andrew Wagenmaker, Sergey Levinecs.LG cs.AI cs.RO stat.ML
In many modern applications of reinforcement learning (RL), the natural reward for a task of interest is inherently sparse: a reward of 0 is given everywhere except when the task is completed, when a reward of +1 is given. Training a policy to maximize such a sparse reward requires solving a challenging credit assignment problem, leading to slow or ineffective RL improvement. We propose a simple approach to transform a sparse outcome reward into a dense process reward. Our approach relies on training a discriminator to distinguish between previous successful and unsuccessful episodes, and using this discriminator to incentivize the RL-learned policy to match the state-action visitations of successful episodes, while avoiding those of unsuccessful episodes. By incentivizing the policy to match the visitations over all states, not just those that correspond to task success, this reward provides dense feedback on whether progress is being made towards task completion, and, we show, provably achieves this without changing the optimal policy. Focusing on finetuning of robotic control policies, we demonstrate that our approach leads to significantly faster RL finetuning performance on both simulated and real-world manipulation tasks, as compared to simply maximizing the sparse outcome reward.
Adversarial imitation learning (AIL) achieves high-quality imitation compared to behavioral cloning (BC), but demands substantial online environment interaction. Recent empirical work has explored initializing AIL algorithms with BC pretrained policies to address this limitation, yet a rigorous theoretical understanding of pretraining's role in AIL remains elusive. This paper provides a systematic theoretical analysis and introduces principled pretraining algorithms for accelerating AIL. We begin by analyzing AIL with policy pretraining alone, identifying reward error as the dominant source of suboptimality. This reveals a critical and previously overlooked gap: the absence of reward pretraining. Motivated by this finding, we develop a principled policy-reward co-pretraining approach grounded in a reward shaping analysis. Our analysis uncovers a fundamental connection between expert policies and shaping rewards, which naturally gives rise to CoPT-AIL, an approach that jointly pretrains both policy and reward through a single BC procedure. We prove that CoPT-AIL achieves an improved imitation gap bound over standard AIL, establishing the first theoretical guarantee for the benefits of pretraining in AIL. Experimental results confirm CoPT-AIL's superior performance over existing AIL methods.
Despite its massive player base and complex hidden-information dynamics, Indian Rummy has received no reinforcement learning attention. Existing agents rely on combinatorial search, which is tactically strong but slow at inference. We present IRumAI, the first RL agent for the domain. IRumAI integrates Proximal Policy Optimization (PPO), meld-aware observation encoding, deadwood-driven reward shaping, and a dual-branch convolutional architecture. IRumAI is RL-trained solely against weak heuristics, after a one-time behaviour-cloning warm-start on stronger demonstration data. It generalises to defeat the entire baseline hierarchy, including a 53.9% win rate against the strongest search-based opponent unseen during RL training. Bypassing explicit search, IRumAI requires just 0.33 ms per action, which is over 7,000x faster than the state-of-the-art heuristic. Ablations validate our architectural choices, and linear probing reveals that the network implicitly models the opponent's hidden hand from public interactions.
The temporal structure of reward composition in reinforcement learning (RL) is typically hand-designed and held fixed throughout training, leaving the progression of motivational priorities largely unexplored. In this work, we propose an evolutionary framework for discovering developmental reward schedules, in which three distinct biologically inspired motivational components -- agency, novelty, and reactivity -- are combined through time-varying weights that dynamically shift over the course of training. Evaluated on two sparse-reward MiniGrid tasks: DoorKey-6x6 and KeyCorridorS3R1, our framework compares the generalizability of four evolutionary algorithms: CMA-ES, xNES, DE, and L-SHADE against an extrinsically motivated baseline (our main comparison point), and three additional hand-designed methods. On DoorKey-6x6, all evolved methods outperform the non-evolved baselines, with L-SHADE achieving the best performance -- an approximate relative mean improvement of 11.4% over the extrinsic only baseline. On KeyCorridorS3R1, CMA-ES achieves the best overall performance, with the remaining evolved methods showing weaker and less reliable generalization capability compared to the extrinsic only baseline. Interestingly, the discovered schedules diverge from our defined developmental ordering, with novelty consistently emerging as the dominant early signal during training, across both tasks. Collectively, our results position evolutionary optimization as a promising approach for developmental reward schedule discovery in deep reinforcement learning, and suggest that what evolution finds to be optimal in computational settings may differ from what it finds to be optimal in biology. The code for this project can be found at: https://github.com/alannadels/Evolutionary_RL.git.