This reasoning-action loop continues until task completion, enabling the VLM to focus on semantic reasoning while delegating precise execution to robust controllers. The framework is evaluated in simulation and real-world experiments using a pretrained VLM, and comprehensive comparison and ablation studies are carried out to verify its performance. CLIPSeg [12] is used for prompt-based segmentation, maintaining a unified prompting pipeline from perception to reasoning. A. Additional Related W orks Aerial manipulation has progressed from vision-guided approaches relying on onboard cameras and artificial visual cues [13], to fully markerless grasping systems using onboard perception [14], and more recently end-effector-centric frameworks for versatile manipulation [15], yet all remain focused on execution rather than language-level reasoning. In parallel, VLAs [2]-[5] combine LLMbased planning [16], [17] with perceptual grounding from models such as CLIP [18], CLIPort [19], and LLaV A [20], but their end-to-end policies are data-intensive and prone to unsafe behaviors from ambiguous outputs, or adversarial prompts, motivating hybrid approaches where reasoning is decoupled from execution via modular skill primitives [21], [22]. For multirotors specifically, foundation model research has focused on mission planning [23], spatial reasoning [24], and direct control [25] which advances locomotion but does not extend to aerial manipulation, and it requires exploration coupled with grasping and placement [26]. In summary, control-focused aerial manipulation, reasoning-focused VLAs, and navigation-focused UA V -VLN each address parts of the problem, but none unify perception, reasoning, and execution for aerial manipulation. Together, these limitations motivate AERMANI-VLM, which unifies open-vocabulary perception, structured reasoning, and safe skill execution for aerial manipulation.

The performance of multi-agent reinforcement learning (MARL) in partially observable environments depends on effectively aggregating information from observations, communications, and reward signals. While most existing multi-agent systems primarily rely on rewards as the only feedback for policy training, our research shows that introducing auxiliary predictive tasks can significantly enhance learning efficiency and stability. We propose Belief-based Predictive Auxiliary Learning (BEPAL), a framework that incorporates auxiliary training objectives to support policy optimization. BEPAL follows the centralized training with decentralized execution paradigm. Each agent learns a belief model that predicts unobservable state information, such as other agents' rewards or motion directions, alongside its policy model. By enriching hidden state representations with information that does not directly contribute to immediate reward maximization, this auxiliary learning process stabilizes MARL training and improves overall performance. We evaluate BEPAL in the predator-prey environment and Google Research Football, where it achieves an average improvement of about 16 percent in performance metrics and demonstrates more stable convergence compared to baseline methods.

Most existing software lacks accessible Application Programming Interfaces (APIs), requiring agents to operate solely through pixel-based Graphical User Interfaces (GUIs). In this API-free setting, large language model (LLM)-based agents face severe efficiency bottlenecks: limited to local visual experiences, they make myopic decisions and rely on inefficient trial-and-error, hindering both skill acquisition and long-term planning. To address these challenges, we propose KG-Agent, an experience-driven learning framework that structures an agent's raw pixel-level interactions into a persistent State-Action Knowledge Graph (SA-KG). KG-Agent overcomes inefficient exploration by linking functionally similar but visually distinct GUI states, forming a rich neighborhood of experience that enables the agent to generalize from a diverse set of historical strategies. To support long-horizon reasoning, we design a hybrid intrinsic reward mechanism based on the graph topology, combining a state value reward for exploiting known high-value pathways with a novelty reward that encourages targeted exploration. This approach decouples strategic planning from pure discovery, allowing the agent to effectively value setup actions with delayed gratification. We evaluate KG-Agent in two complex, open-ended GUI-based decision-making environments (Civilization V and Slay the Spire), demonstrating significant improvements in exploration efficiency and strategic depth over the state-of-the-art methods.

The bisimulation metric (BSM) is a powerful tool for computing state similarities within a Markov decision process (MDP), revealing that states closer in BSM have more similar optimal value functions. While BSM has been successfully utilized in reinforcement learning (RL) for tasks like state representation learning and policy exploration, its application to multiple-MDP scenarios, such as policy transfer, remains challenging. Prior work has attempted to generalize BSM to pairs of MDPs, but a lack of rigorous analysis of its mathematical properties has limited further theoretical progress. In this work, we formally establish a generalized bisimulation metric (GBSM) between pairs of MDPs, which is rigorously proven with the three fundamental properties: GBSM symmetry, inter-MDP triangle inequality, and the distance bound on identical state spaces. Leveraging these properties, we theoretically analyse policy transfer, state aggregation, and sampling-based estimation in MDPs, obtaining explicit bounds that are strictly tighter than those derived from the standard BSM. Additionally, GBSM provides a closed-form sample complexity for estimation, improving upon existing asymptotic results based on BSM. Numerical results validate our theoretical findings and demonstrate the effectiveness of GBSM in multi-MDP scenarios.

Modelling pedestrian-driver interactions is critical for understanding human road user behaviour and developing safe autonomous vehicle systems. Existing approaches often rely on rule-based logic, game-theoretic models, or 'black-box' machine learning methods. However, these models typically lack flexibility or overlook the underlying mechanisms, such as sensory and motor constraints, which shape how pedestrians and drivers perceive and act in interactive scenarios. In this study, we propose a multi-agent reinforcement learning (RL) framework that integrates both visual and motor constraints of pedestrian and driver agents. Using a real-world dataset from an unsignalised pedestrian crossing, we evaluate four model variants, one without constraints, two with either motor or visual constraints, and one with both, across behavioural metrics of interaction realism. Results show that the combined model with both visual and motor constraints performs best. Motor constraints lead to smoother movements that resemble human speed adjustments during crossing interactions. The addition of visual constraints introduces perceptual uncertainty and field-of-view limitations, leading the agents to exhibit more cautious and variable behaviour, such as less abrupt deceleration. In this data-limited setting, our model outperforms a supervised behavioural cloning model, demonstrating that our approach can be effective without large training datasets. Finally, our framework accounts for individual differences by modelling parameters controlling the human constraints as population-level distributions, a perspective that has not been explored in previous work on pedestrian-vehicle interaction modelling. Overall, our work demonstrates that multi-agent RL with human constraints is a promising modelling approach for simulating realistic road user interactions.

Reinforcement learning (RL) has proven effective for fine-tuning large language models (LLMs), significantly enhancing their reasoning abilities in domains such as mathematics and code generation. A crucial factor influencing RL fine-tuning success is the training curriculum: the order in which training problems are presented. While random curricula serve as common baselines, they remain suboptimal; manually designed curricula often rely heavily on heuristics, and online filtering methods can be computationally prohibitive. To address these limitations, we propose Self-Evolving Curriculum (SEC), an automatic curriculum learning method that learns a curriculum policy concurrently with the RL fine-tuning process. Our approach formulates curriculum selection as a non-stationary Multi-Armed Bandit problem, treating each problem category (e.g., difficulty level or problem type) as an individual arm. We leverage the absolute advantage from policy gradient methods as a proxy measure for immediate learning gain. At each training step, the curriculum policy selects categories to maximize this reward signal and is updated using the TD(0) method. Across three distinct reasoning domains: planning, inductive reasoning, and mathematics, our experiments demonstrate that SEC significantly improves models' reasoning capabilities, enabling better generalization to harder, out-of-distribution test problems. Additionally, our approach achieves better skill balance when fine-tuning simultaneously on multiple reasoning domains. These findings highlight SEC as a promising strategy for RL fine-tuning of LLMs.

Current evaluations of agents remain centered around one-shot task completion, failing to account for the inherently iterative and collaborative nature of many real-world problems, where human goals are often underspecified and evolve. We argue for a shift from building and assessing task completion agents to developing collaborative agents, assessed not only by the quality of their final outputs but by how well they engage with and enhance human effort throughout the problem-solving process. To support this shift, we introduce collaborative effort scaling, a framework that captures how an agent's utility grows with increasing user involvement. Through case studies and simulated evaluations, we show that state-of-the-art agents often underperform in multi-turn, real-world scenarios, revealing a missing ingredient in agent design: the ability to sustain engagement and scaffold user understanding. Collaborative effort scaling offers a lens for diagnosing agent behavior and guiding development toward more effective interactions.

In Vision-Language-Action (VLA) models, robustness to real-world perturbations is critical for deployment. Existing methods target simple visual disturbances, overlooking the broader multi-modal perturbations that arise in actions, instructions, environments, and observations. Here, we first evaluate the robustness of mainstream VLAs under 17 perturbations across four modalities. We find (1) actions as the most fragile modality, (2) Existing visual-robust VLA do not gain robustness in other modality, and (3) pi0 demonstrates superior robustness with a diffusion-based action head. To build multi-modal robust VLAs, we propose RobustVLA against perturbations in VLA inputs and outputs. For output robustness, we perform offline robust optimization against worst-case action noise that maximizes mismatch in flow matching objective. This can be seen as adversarial training, label smoothing, and outlier penalization. For input robustness, we enforce consistent actions across input variations that preserve task semantics. To account for multiple perturbations, we formulate robustness as a multi-armed bandit problem and apply an upper confidence bound algorithm to automatically identify the most harmful noise. Experiments on LIBERO demonstrate our RobustVLA delivers absolute gains over baselines of 12.6% on the pi0 backbone and 10.4% on the OpenVLA backbone across all 17 perturbations, achieving 50.6x faster inference than existing visual-robust VLAs, and a 10.4% gain under mixed perturbations. Our RobustVLA is particularly effective on real-world FR5 robot with limited demonstrations, showing absolute gains by 65.6% under perturbations of four modalities.

Modeling systems that exhibit both continuous and discontinuous state changes presents a significant challenge in machine learning. For instance, biological spiking networks feature the continuous decay of neuron potentials alongside discontinuous spikes, which cause abrupt increases in the potentials of neighboring downstream neurons. Designing appropriate objective functions and applying gradient methods that work with these discontinuities are among the difficulties of working with such systems. Traditionally, ordinary and partial differential equations (ODEs and PDEs) are used to model continuous state changes, while Markov decision processes (MDPs) are employed to capture discrete actions that drive environmental transitions. In this paper, we study Action-Driven Processes (ADPs), also known as generalized semi-Markov processes [12, 5, 16], which unify both types of dynamics within a single framework. With continuous-time states and actions at the core of ADPs, a natural question is whether it is possible to learn optimal policies for action selection using traditional reinforcement learning methods. The control-as-inference tutorial [9] elegantly demonstrated that maximum entropy reinforcement learning can be formulated as minimizing the Kullback-Leibler (KL) divergence between (a) a true trajectory distribution generated by action-state transitions and the policy, and (b) a model trajectory distribution that depends on the reward function.

Privacy concerns have become increasingly critical in modern AI and data science applications, where sensitive information is collected, analyzed, and shared across diverse domains such as healthcare, finance, and mobility. While prior research has focused on protecting privacy in a single data release, many real-world systems operate under sequential or continuous data publishing, where the same or related data are released over time. Such sequential disclosures introduce new vulnerabilities, as temporal correlations across releases may enable adversaries to infer sensitive information that remains hidden in any individual release. In this paper, we investigate whether an attacker can compromise privacy in sequential data releases by exploiting dependencies between consecutive publications, even when each individual release satisfies standard privacy guarantees. To this end, we propose a novel attack model that captures these sequential dependencies by integrating a Hidden Markov Model with a reinforcement learning-based bi-directional inference mechanism. This enables the attacker to leverage both earlier and later observations in the sequence to infer private information. We instantiate our framework in the context of trajectory data, demonstrating how an adversary can recover sensitive locations from sequential mobility datasets. Extensive experiments on Geolife, Porto Taxi, and SynMob datasets show that our model consistently outperforms baseline approaches that treat each release independently. The results reveal a fundamental privacy risk inherent to sequential data publishing, where individually protected releases can collectively leak sensitive information when analyzed temporally. These findings underscore the need for new privacy-preserving frameworks that explicitly model temporal dependencies, such as time-aware differential privacy or sequential data obfuscation strategies.