Search research

We present HERCULES, an open-source simulator and data-collection pipeline for heterogeneous multi-robot autonomy. Built upon the Unreal Engine 5 (UE5)-based simulators AirSim and Cosys-AirSim, HERCULES resolves key architectural limitations of prior frameworks to enable concurrent unmanned aerial and ground vehicle (UAV-UGV) operation in large-scale, photorealistic, dynamic environments. It introduces a new waypoint-tracking UGV controller that mirrors existing UAV control interfaces, and provides a shared navigation stack for mapping, traversability analysis, planning, and control across heterogeneous platforms. Expanding inherited sensor suites, it adds physics-based long-wave infrared (LWIR) cameras and configurable night-vision modes for degraded visual environments. HERCULES provides lightweight APIs, ROS 2 wrappers, and rigorous time synchronization across sensors and platforms, and brings state-of-the-art game-engine capabilities into robotics simulation, integrating intelligent agents such as pedestrians, traffic, and wildlife with high-fidelity dynamic phenomena, including fire, flooding, and crop disease spread. HERCULES runs in two modes: passively, replaying offline-designed trajectories to generate reproducible multi-modal datasets, and actively, running an online planner in closed loop from live observations. Our experiments in heterogeneous multi-robot SLAM, collaborative perception, and exploration, using both HERCULES-generated data and active closed-loop execution, demonstrate its utility for advancing heterogeneous multi-robot autonomy. We publicly release our source code, experiment code, documentation, and datasets, including a heterogeneous multi-robot SLAM benchmark collected with two UAVs and two UGVs across kilometer-scale desert, forest, and city environments, at https://lunarlab-gatech.github.io/HERCULES-website.

The imperative to develop more sustainable agriculture demands a transition from isolated automation toward the deployment of multi-robot systems (MRS) in agrifood environments. However, Mediterranean greenhouse settings-characterized by narrow corridors, dense biomass, and structural metallic interference-pose significant challenges for robust and scalable communication between agents. Traditional robotic frameworks, such as ROS 2, frequently encounter node discovery issues and latency spikes due to dynamic obstacles, dense foliage, and other characteristic greenhouse elements, creating a critical bottleneck for real-time coordination. This paper proposes an innovative cloud-based hybrid architecture that establishes a two-way communication bridge between ROS 2, acting as an edge computing platform, and iVeg as a Decision Support System (DSS), using MQTT and the European FIWARE platform. The proposed framework enables seamless interoperability between fleets of multiple robots in environments with communication constraints, facilitating the synchronised exchange of high-level telemetry, point cloud data and farmer identification for collaboration, amongst other critical data. The architecture was validated in a high-fidelity simulation environment and subsequently tested in a real-world greenhouse scenario, demonstrating its ability to maintain persistent connectivity and data integrity under adverse network conditions. The results indicate that the integration of MQTT effectively eliminates information silos, providing a scalable and decentralised solution for managing complex robotic missions, which are executed locally via Edge Computing. This work sets a new methodological precedent for the concept of Greenhouse Models as a Service (GMaaS), bridging the gap between low-level robotic control and high-level, cloud-based IoT decision-making.

Jamming and spoofing pose significant threats to wireless and satellite navigation by disrupting radio-frequency (RF) signals and compromising availability and integrity. Robust RF interference direction finding through angle-of-arrival (AoA) estimation is therefore essential for detecting and localizing anomalous signals. Although data-driven methods perform well under line-of-sight (LoS) conditions, their performance degrades in practical environments due to non-line-of-sight (NLoS) multipath propagation. In this work, we propose a hybrid learning framework that incorporates physics-informed constraints into deep neural networks to improve the robustness of AoA estimation. A neural network is trained to estimate the azimuth and elevation of incoming signals received by a four-element antenna array, while a physics-informed loss enforces consistency between the predicted angles and inter-antenna phase differences under a plane-wave model. We further introduce a latent-space classifier to distinguish LoS from NLoS samples. Since inter-antenna phase differences under LoS propagation exhibit domain-invariant structure across environments, the physics-based loss is applied only to LoS samples, promoting physically consistent and domain-invariant representations without over-constraining the model in NLoS scenarios. In addition, domain-incremental learning (DIL) across NLoS environments with varying scatterer distributions improves cross-domain generalization. Evaluations on real-world datasets show that the proposed method reduces AoA estimation error by up to 6° in low-exemplar settings compared with DIL baselines.

Model-free reinforcement learning algorithms such as Proximal Policy Optimization (PPO) treat the environment as a black box, estimating policy gradients from sampled rewards; this process demands millions of interactions and relies on high-variance advantage estimates. When environment dynamics are differentiable, the return is an end-to-end differentiable function of the policy parameters, enabling exact gradient computation via backpropagation through simulation. We term this approach Analytic Policy Gradients (APG) and evaluate it against PPO on four continuous control tasks of increasing dynamical complexity: a one-dimensional point-mass target-reaching task, a 2D point-mass navigation task with obstacle avoidance, a 2D rigid-body T-block pushing task, and a 7-DOF Franka FR3 end-effector reaching task. Both algorithms share identical model architectures, observation normalization, and optimizer settings. To decouple sample efficiency from compute efficiency, we design a multi-axis evaluation protocol that records performance against environment steps and gradient steps. We report a segmented backpropagation scheme with MC and critic-based bootstrap modes that mitigates gradient degradation on long-horizon tasks, and present ablations over segment length and bootstrap strategy.

Physically localizing unauthorized wireless devices remains a critical bottleneck in cyber-physical security operations, where rogue access points can provide entry points for lateral movement and persistent compromise. While such devices can often be detected through network-side mechanisms, determining their physical location typically requires dense sensing infrastructure, site-specific RF fingerprinting, or manual inspection, limiting timely incident response. We investigate whether a single commodity robot can autonomously detect and localize rogue wireless devices under zero-configuration constraints, without RF fingerprinting, pre-installed sensors, or site calibration. We present RogueRover, an end-to-end system in which a quadruped robot autonomously patrols, collects spatially labeled RSSI measurements via a standard 802.11 interface, and estimates device locations offline. We evaluate the system across 11 patrol runs in a real indoor environment, with 6 rogue devices deployed under heterogeneous propagation conditions. Across 62 AP-patrol sessions, RogueRover achieves a median single-patrol localization error of 1.62 m without prior RF knowledge. Under multi-run aggregation, five of six devices are localized within 1 m. A blind trial validates the full pipeline, correctly identifying rogue devices among 73 observed BSSIDs and localizing them with errors of 0.34 m and 1.84 m. Across environments, simple weighted-centroid estimators perform comparably to, or better than, parametric path-loss models, indicating that measurement coverage from autonomous patrols is the primary determinant of localization accuracy under zero-prior constraints. Our results demonstrate that infrastructure-free, autonomous localization is feasible in practice, enabling rapid physical incident response in cyber-physical environments without additional sensing infrastructure.

Deploying highly capable personalized conversational agents in resource-constrained or privacy-sensitive environments remains a significant challenge. We identify a fundamental bottleneck in the existing approaches: current training paradigms treat personalization and grounding as a single monolithic learning problem. Under these paradigms, language models are forced to simultaneously address what to say (content grounding) and how to say it in a user-specific way (personalization), which introduces significant computational and optimization challenges. Consequently, contextual grounding is often sacrificed for persona adherence, or vice versa, resulting in responses that are either weakly grounded in the conversational history or insufficiently personalized. In this work, we propose the Generic Response-Augmented Generation (GRAG) framework that decouples these competing objectives by leveraging offline, generic responses from high-capacity, general-purpose LLMs as a semantic and structural scaffold to guide the fine-tuning of smaller, task-specialized models seamlessly in resource-limited environments. By decoupling the content grounding from personalization, GRAG allows the model to focus exclusively on persona injection while remaining firmly anchored to the conversational context. We instantiate the GRAG in two post- and pre-fusion-based architectural variants and evaluate them on multiple benchmark conversational datasets that cover diverse personalization structures. Our results demonstrate that GRAG significantly outperforms state-of-the-art methods that do not use auxiliary scaffolding, yielding up to 47% improvements in ROUGE-2 and 36% in BLEU scores. Ultimately, GRAG offers a generalizable blueprint for building grounding-aware personalized conversational systems in resource-limited environments.

Many decision-making problems in computing and networking systems can be naturally formulated as cost-minimization problems under performance constraints. In dynamic environments, reinforcement learning (RL) is often used to solve such problems at runtime by embedding both costs and constraint violations into a single scalar reward through weighted penalty terms, following a Lagrangian-inspired formulation. However, in this context the behavior of the learned policy critically depends on the choice of these weights, which are typically selected manually. This makes it difficult to identify an appropriate trade-off between optimizing the primary objective and effectively avoiding constraint violations, particularly in non-stationary environments where their relative importance may change. This paper presents MAMO (Multi-Agent system for Multi-Objective constrained optimization), an approach to tackle this balancing problem through multi-agent RL. MAMO decouples task execution from objective design by formulating the selection of reward weights as a learning problem, providing a !rst step towards more autonomous and robust RL-based solutions for constrained optimization problems in dynamic environments.

Language model agents are becoming proficient executors at isolated, short-horizon tasks such as software engineering and customer service. Yet real-world challenges require a combination of sophisticated skills that remain largely untested in agents: (1) navigating long horizons amid uncertainty; (2) acquiring information in noisy environments; (3) adapting to a changing world; (4) orchestrating multiple moving parts toward a coherent goal. We introduce CEO-Bench, which evaluates these capabilities together by simulating a representative real-world task: operating a startup for 500 days. An agent manages pricing, marketing, budgeting, and many other aspects of a fictional company through a programmable Python interface, operating in the same environment and facing the same challenges as a human CEO. Success demands analyzing noisy, interconnected business databases, translating signals into sound strategy, and coordinating many decisions with programming. The strongest agents write sophisticated code that simulates customer cohorts to forecast future cash and mines negotiation history to uncover hidden customer preferences. Even so, most state-of-the-art models struggle in this environment. Only Claude Opus 4.8 and GPT-5.5 finish above the $1M starting balance, and neither consistently turns a profit. CEO-Bench takes a first step toward measuring the intelligence required to drive sustained, adaptive progress over time.

Robot learning and embodied agents now require simulation to serve as a shared execution substrate linking control, skills, and planning, not only as a renderer, controller testbed, or fixed task environment. Existing pipelines split these layers with "magic" actions, disconnected training environments, or forward-only renders that cannot reproduce, evaluate, and annotate the same episode. We present MagicSim, an embodied interaction infrastructure built around one deterministic batched runtime and a shared Markov decision process (MDP). From YAML-first specifications that decouple contents, placement, behavior, and agent exposure, MagicSim constructs diverse executable worlds spanning task families, interaction regimes, physics, layouts, sensors, avatars, and robot embodiments in one reset-and-step loop. A common execution interface grounds high-level commands through controllers, atomicskills, planner primitives, and asynchronous planning, realizing them as robot actions rather than simulator-side state edits. One task definition supports three capabilities: benchmark and RL evaluation, an autocollect interface that automatically turns commands into grounded trajectories, and agent/VLM-facing interaction. For automatic execution, commands flow through a Command->Skill->Planner->Robot->Record pipeline, while per-environment command, skill, planning, retry, annotation, and episode states advance independently above the shared physics tick. Successful rollouts are saved as structured multimodal trajectories aligning language supervision, action representations, visual/geometric representations, and task-level status with the executed episode. MagicSim thus unifies diverse world construction, embodied execution, task evaluation, automatic rollout generation, and interactive agent interfaces in one planner-in-the-loop runtime.

Out-of-distribution (OOD) detection in dynamic open-world environments requires a model to continually adapt to evolving data distributions while generalizing to covariate-shifted inputs and rejecting semantic-shifted OOD examples. Most existing OOD detection methods optimize only the current-step objective and do not explicitly account for how post-deployment environment changes affect future OOD behavior. In this paper, we establish a theoretical grounding for dynamic OOD detection using a reinforcement learning (RL)-guided optimizer that explicitly favors updates that reduce the semantic OOD false positive rate over time. We develop a novel augmented optimizer that uses an RL-guided correction term on top of standard gradient descent (GD) and show its improvement over both future-domain generalization and semantic-OOD rejection. We analyze temporal error decomposition in terms of model-change and environment-change generalization errors and develop a new theoretical framework for comparing the generalization errors under both GD and RL-guided optimizers.

Robotic assistance in household environments requires not only predicting where objects should be placed, but also reasoning about when objects should not be placed at all. Existing approaches to personalized object rearrangement primarily focus on placement decisions under the assumption of clean observations and complete actionability, limiting their applicability in realistic, cluttered, and partially erroneous settings. In this paper, we introduce APOLLO, a hybrid framework for abstention-aware personalized object rearrangement that combines a lightweight, personalized embedding model (PEM) with selective large language model (LLM) assistance. PEM is trained for each user-environment pair using a small number of demonstrations, operates entirely on CPU, and produces uncertainty estimates, which are used to selectively invoke LLM-based reasoning only for ambiguous decisions, balancing efficiency, privacy, and reasoning capability. To evaluate this formulation beyond existing benchmarks, we introduce APOR, a synthetic, LLM-generated dataset that captures room-level, multi-furniture environments, diverse organizational profiles, explicit abstention behavior, and noisy partial scene context. Extensive experiments on both PARSEC and APOR provide initial evidence that APOLLO improves over prior LLM-based baselines in controlled benchmark settings while substantially reducing LLM usage. Code is available at https://github.com/PaInt-Lab/APOLLO.

Reinforcement learning (RL) systems often degrade when operating conditions differ from those previously encountered, reflecting distributional shifts in the underlying data-generating process. Such shifts may occur between training and evaluation, as in In-Distribution (ID) and Out-of-Distribution (OOD) generalization, or within non-stationary settings where environment dynamics evolve over time. However, the formal relationship between these views remains unclear, and existing work mainly focuses on mitigation rather than the causal origin of shift within the agent-environment interaction. This work develops a unified causal-origin taxonomy that characterizes sources of distributional shift in RL and relates ID/OOD generalization to non-stationary settings. We transfer the classical dataset-shift principle from supervised learning to RL by reformulating distributional shift in terms of the generative interaction process. Using a Partially Observable Markov Decision Process (POMDP), we decompose the interaction into structural components, including the state distribution, observation process, policy, reward, and transition dynamics, together with the shifted-time boundary. The proposed taxonomy distinguishes internal (agent-driven) and external (environment-driven) distributional shifts. The shifted-time boundary perspective further characterizes explicit, implicit, and hybrid shifts. This formulation unifies ID/OOD generalization and non-stationarity as structured changes in the underlying process. We also introduce an evaluation framework for measuring shift impact and adaptation through performance degradation and recovery metrics. By grounding distributional shift in the causal-origin structure of RL, this work supports systematic analysis of robustness under distributional shift.

Current benchmarks for computer-use agents evaluate models in impersonal environments. This leaves a gap between evaluation and deployment where personal assistants are expected to work across a user's whole digital life, including their context, historical data, and logged-in accounts. This gap is widest on web tasks, where live web evaluations cannot exercise sites that require logging in or personal information, the kind of site a real personal assistant has to drive. We introduce MyPCBench, which tests computer-use agents as personal assistants on a Linux desktop populated with 17 simulated real-world web applications and a full desktop stack, all seeded for one canonical persona, Michael Scott from The Office. We define 184 tasks in this environment, each inspired by a real request drawn from the OpenClaw community, and benchmark six closed and open-weight models with a uniform computer+bash tool surface. We find that the best model, Claude Opus 4.6, fully solves 55.4% of the tasks, the only model above 50%. Model failures cluster on tasks that span many applications and on long trajectories, where personalization stresses an assistant the most. We release the environment, task set, and agent harness at https://mypcbench.com.

User instructions are often underspecified because humans rely on implicit assumptions about the surrounding environment. For large language model (LLM) agents operating in information-rich digital and physical environments, these assumptions cannot be inferred from the instruction alone; they must be recovered from the current state of tools, data, interfaces, and observations. Effective execution therefore requires agents to identify missing context, ground it in observed evidence, and carry it forward into subsequent actions. We show that current agents often fail to do so. They act from assumed rather than observed specifics, overlook information they could have gathered, and fail to incorporate evidence that has already been returned. Building on this insight, we propose ACCORD (Action-Conditioned Contextual Grounding), a simple and effective agent framework for adaptive grounding. Before each action, ACCORD actively probes the environment for missing information and integrates relevant context from the agent's trajectory that would otherwise be overlooked. Requiring no additional training or task-success signals, ACCORD improves task-goal completion on AppWorld by up to +20.6 points with GPT-5-mini, from 42.0% to 62.6%, compared to strong baselines. These gains persist with a substantially stronger base model (+10.8 with Claude-4.5-sonnet), an open-weight model (+10.1 with Qwen3.5-27B-FP8), and on the embodied AlfWorld benchmark (+7.4 success rate with GPT-5-mini).

We present TickingCollabBench, a Minecraft-based multi-agent benchmark for a novel class of time-sensitive complementary collaboration tasks. Our benchmark reflects four core characteristics of real-world collaboration: agent heterogeneity, mandatory collaboration, dynamic environments, and strict real-time constraints with failure risks. To enable this, we develop the TickingCollab framework, which supports the generation of diverse dynamic environments and abstracts Minecraft's primitive APIs to enable declarative YAML task specifications for composing these events. Building on this, we design a feasibility-aware automated benchmark generation pipeline, where an LLM drafts structurally diverse task configurations and feasibility verifier filters out invalid ones using approximate constraints. Evaluations demonstrate that lang latency and inherent difficulty of coordinating under partial observability and agent heterogeneity cause LLMs to frequently fail under dynamic environments and fall significantly short of a global-knowledge oracle.

The success of generative models in language and visual generation has inspired extensive applications to generative robot planning. However, most existing works either focus on single-robot planning, or generate multi-robot trajectories in a sequential manner with iterative post-processing to resolve inter-robot conflicts. In this work, we investigate whether coordinated multi-robot trajectories, as a special spatiotemporal distribution, can be learned and generated with a generative model in a feed-forward manner. We propose Robots as Tokens (Roken), a unified diffusion transformer that directly generates multi-robot trajectories that satisfy both (individual) safety and (global) connectivity constraints. The core design of Roken is to represent each robot as a discrete token, allowing them to naturally interact with each other through self-attention, and cross-attend to map tokens for environment layouts. We further introduce several auxiliary tasks based on Bayes' theorem to provide multi-scale spatial-temporal supervision for efficient learning of the conditional distribution. In training, Roken absorbs diverse expert trajectories from different team sizes. During inference, Roken behaves as a versatile multi-robot planner that can handle single-robot planning, coordinated multi-robot trajectory generation, and conditional trajectory generation by fixing some robot tokens as conditions. Experiments in diverse cluttered environments show that Roken can generate coordinated multi-robot trajectories to perform connectivity-constrained goal navigation tasks with high success rates, outperforming the baseline method used to generate the training dataset. Roken also demonstrates good scalability after training with mixed team sizes, and shows generalization to unseen or partially observed environments, verifying its potential to learn from diverse data and perform versatile tasks.

Advanced agents are increasingly demonstrating the potential to operate as autonomous engineers, creating a growing demand for evaluation benchmarks that capture the complexity of real-world development. Such environments typically involve both complex code and large-scale data (i.e., file system). However, existing benchmarks usually evaluate code-centric or data-centric capabilities in isolation, leaving a clear gap with real development scenarios. In this paper, we bridge this gap by introducing CODA-BENCH, the first benchmark to jointly evaluate code and data intelligence in a data-intensive environment. We construct a data-intensive Linux sandbox based on the Kaggle ecosystem (containing hundreds of datasets), where agents must actively explore complex file hierarchies to identify relevant resources and generate code for data-driven analytical tasks. CODA-BENCH comprises 1,009 tasks spanning 31 communities, with each task environment containing an average of 980 files, simulating realistic data scale and noise. Evaluations of advanced agents reveal that even top-performing systems struggle to effectively integrate data discovery with code execution, achieving a success rate of only 61.1%. These results highlight a substantial gap in current agentic capabilities for data-intensive tasks and point to promising directions for future research.

Efficient and scalable agentic intelligence requires models that can deliver both low-latency responses and strong reasoning capabilities while remaining practical to train, serve, and deploy. In this report, we present Ling-2.6 and Ring-2.6, a family of models designed to address this challenge at scale. Ling-2.6 is optimized for instant response generation and high capability per output token, whereas Ring-2.6 is tailored for deeper reasoning and more advanced agentic workflows. Instead of training from scratch, we upgrade the Ling-2.0 base model through architectural migration pre-training and large-scale post-training. This upgrade is guided by a unified co-design of model architecture, optimization objectives, serving systems, and agent training environments, enabling improvements in both model capability and deployment efficiency. At the architectural level, we introduce a hybrid linear attention design that integrates Lightning Attention with MLA, improving the efficiency of long-context training and decoding. To further enhance token efficiency, we optimize capability per output token through Evolutionary Chain-of-Thought, Linguistic Unit Policy Optimization, bidirectional preference alignment, and shortest-correct-response distillation. For agentic capabilities, we propose KPop, a reinforcement learning framework designed to support stable training of Ring-2.6-1T on large-scale environment-grounded data. KPop improves training efficiency through asynchronous scheduling across coding, search, tool use, and workflow execution, enabling scalable learning from complex agent-environment interactions. Together, Ling-2.6 and Ring-2.6 provide a practical pathway toward efficient, scalable, and open agentic systems. We open-source all checkpoints in the 2.6 family to support further research and development in practical agentic intelligence.

Mobile agents require efficient exploration strategies to map unseen environments and autonomously plan tasks. Traditional methods rely on generating occupancy maps and optimizing the sequence in which unexplored regions are visited. However, in sensor-constrained settings, such as those limited to monocular cameras, generating accurate occupancy maps is challenging. To address this, we propose VANDERER, an exploration framework that leverages a Visual Curiosity Module (VCM) to guide pre-trained diffusion policies using only monocular image data. This curiosity module predicts the outcomes of proposed actions via a navigation world model and evaluates them through a curiosity cost. The cost then guides the diffusion process toward generating actions that maximize exploration. Evaluated across diverse simulated environments, VANDERER consistently outperforms established baselines, exploring an average of 13.4% more area than NoMaD. Our results reveal a direct correlation between visual and geometric curiosity in outdoor environments, demonstrating that VANDERER can effectively leverage this relationship for efficient exploration using sensor-constrained agents.

LLMs are increasingly deployed as agents that interact with external environments and observe feedback such as execution results, error messages, and tool outputs. A well-functioning agent should be able to leverage this feedback to accurately assess its own performance. Yet we find a persistent reflection gap: LLM agents tend to mis-assess their own outputs after observing concrete environment feedback -- even for questions they correctly answered -- and standard RL barely helps due to a credit-assignment mismatch. To close this gap, we propose RefGRPO, a simple yet effective fix that augments standard RL algorithms with two key ingredients: a free calibration bonus computed by contrasting the agent's own reflection with the actual outcome (requiring no additional reward model, LLM judge, or external annotation), and a dynamic schedule on its coefficient. Compared to standard RL baselines, our method simultaneously improves reflection calibration (e.g., reduces underconfidence rate $44.4\% \to 7.7\%$) and task accuracy (e.g., $75.1\% \to 76.5\%$) on text-to-SQL across five benchmarks. The resulting calibrated reflection turns the agent into its own verifier grounded in environment feedback, which further enables (i) better self-improvement that uses reflections as pseudo-rewards without outcome supervision, and (ii) more effective test-time selective prediction by committing only to rollouts flagged as correct.

As a long-standing challenge in robotic manipulation, stable and efficient grasping in cluttered environments is of great importance in industrial settings. While recent studies have achieved relatively high success rates in grasping from clutter, there remain few mature solutions for more demanding tasks such as sequential object search and sorting. This work addresses sequential object picking in cluttered environments based on the Cluttered Environment Picking Benchmark (CEPB) and presents our solution to the Pick-in-Clutter track of the 10th Robotic Grasping and Manipulation Competition (RGMC) at ICRA 2025. The task poses several key challenges. First, it requires robust and collision-aware grasping with high success rates across a diverse set of objects, including both rigid and deformable ones. Second, it demands efficient search for target objects, which places stringent requirements on the decluttering and searching strategies of the solution. To address the above challenges, we design an integrated hardware-software pipeline that combines object recognition, decluttering, and multi-modal grasping. The main contributions include the hardware design of a multifunctional gripper and novel representations for object distribution and occlusion relationships in cluttered space. This pipeline enables efficient recognition, search, and sequential grasping of objects in clutter, demonstrating strong performance in both laboratory tests and competition scenarios, and ultimately achieving second place in the Pick-in-Clutter track of the RGMC 2025.

Large language models are increasingly deployed as agents for long-horizon tasks, yet their performance is shaped not only by model capability and environment design, but also by the harness that mediates agent--environment interaction. Existing harnesses are largely manually engineered, making them difficult to scale as trajectories grow longer and interactions become more complex. In this work, we ask whether harness can be generated by a learnable plug-in module that can be trained in an end-to-end fashion. We introduce HarnessBridge, a lightweight learnable harness controller that parameterizes the agent--environment interface as a bidirectional projection. HarnessBridge learns two bidirectional projections: observation projection, which distills raw trajectories into compact, decision-relevant states, and action projection, which converts proposed actions into executable transitions or trajectory-grounded rejections. We train HarnessBridge on a harness supervision dataset via unified instruction tuning. On Terminal-Bench~2.0 and SWE-bench Verified, HarnessBridge matches or surpasses strong specialized harnesses while substantially reducing token usage and trajectory length, and generalizes from smaller generators to larger commercial models.

Advanced AI systems have extensive knowledge of their environments; in fact, their knowledge may (far) exceed that of their developers or users. Consequently, a desirable property for an AI system is that it is honest -- that it accurately reports its beliefs about the world. Designing an AI system to be honest may be difficult, especially if we want to ask it questions about latent variables in the environment -- variables which are hidden from the human interacting with it. This gives rise to the problem of eliciting latent knowledge (ELK): the problem of training an AI agent to honestly report its beliefs. In this paper, we make ELK formally precise using Causal Influence Diagrams (CIDs). CIDs can be used to describe the relationship between an agent's training environment and its subjective representation of the world. We use CIDs to formalise the distinction between observable and latent variables, to specify what exactly it means for an agent to be honest, and to formally define goal misgeneralisation. We show that, under certain circumstances, developers can incentivise an agent to honestly answer questions by providing correct feedback during training. However, a natural, but undesirable, way for an agent to generalise is to provide answers which humans would evaluate as true, rather than honest answers. We prove an impossibility theorem stating: There is no feedback-based training strategy that depends only on agent behaviour and with certainty produces an honest agent, even if feedback is perfect during training.

Deep research and agent evolution serve as de-facto tasks for AI agents in real-world applications toward artificial general intelligence. The former enables autonomous retrieval and integration of information in open-ended environments to tackle open-ended research tasks, yet it is constrained by the static parametric deep research capabilities of agent systems. The latter allows agents to autonomously interact with the environment to gain experiences that evolve model capabilities. However, its effectiveness has been widely validated only on verifiable tasks with standard answers, leaving a gap with open-ended research tasks. To bridge these two critical tasks, we propose the Hybrid Open-Ended Tri-Evolution (HOTE) framework, which leverages hybrid-mode reinforcement learning to facilitate the collaborative evolution of a proposer, solver and judge based on web-scale knowledge, moving toward autonomous evolving agents in open-ended tasks and environments. Extensive experiments on three long-form deep research benchmarks demonstrate that the 8B model trained via HOTE surpasses the strongest static open 8-32B models as well as those trained by state-of-the-art deep research training methods with less time overhead, and further verify that the evolution of all three modules in HOTE is indispensable.

Studies on rodents such as mice have shown the capabilities to adapt their behavior when dealing with changing parameters (drift'') of the environment even if no information about change is provided (uncertainty) -- a behavior that can be modeled by forgetting mechanisms. Non-stationary Reinforcement Learning (NSRL) deals with adapting state-of-the-art RL methods to deal with changing environments: these however usually require (partially) perfect information about the drift such as task IDs'' or ``context''. To mitigate the effects of drift, this work develops \emph{Space-sampled Value Decay} as an explicit forgetting mechanism for value-based deep RL architectures as a simple yet effective approach. In particular we demonstrate and discuss positive effects but also limitations in achieved returns for modifications of Deep Q-networks (DQN) and Soft Actor-Critic (SAC) when evaluated on non-stationary environments.

Recent advances in reasoning and tool-calling capabilities of large language models (LLMs) have enabled increasingly capable agentic systems. However, existing benchmarks remain limited in task complexity, realism, and domain diversity, and often fail to capture interactions that span multiple domains, limiting their ability to evaluate agents in realistic multi-step settings that require sustained reasoning and coordination. To address these limitations, we introduce T1-Bench, a high-fidelity, comprehensive benchmark for evaluating agentic systems in realistic customer-facing, multi-domain environments, featuring interleaved scenarios that require structured reasoning across multi-turn user-assistant interactions and substantially increasing both compositional complexity and evaluative rigor across 25 domains of varying difficulty. We evaluate T1-Bench using 12 proprietary and open-weight models, providing a reproducible and standardized framework for assessing agent behavior, tool utilization, and conversational quality in complex, multi-step environments. We further complement automatic evaluation with human judgments to strengthen the assessment of qualitative performance. Overall, T1-Bench substantially advances prior benchmarks by increasing task complexity, interaction depth, and domain coverage in simulated multi-domain environments. To facilitate future research on agentic systems, we will publicly release data and evaluation code as open source.

Although Large Language Model (LLM) agents have demonstrated strong performance on complex tasks, their learning is often limited by inefficient interaction feedback and static training environments, which hinder broader generalization. To address these limitations, this paper introduces Role-Agent, \textcolor{black}{a framework} that harnesses a single LLM to function concurrently as both the agent and the environment, enabling a bootstrapped co-evolution. Role-Agent comprises two synergistic components: World-In-Agent (WIA) and Agent-In-World (AIW). In WIA, the LLM acts as the agent and predicts future states after each action; the alignment between predicted and actual states is then used as a process reward, encouraging environment-aware reasoning. In AIW, the LLM analyzes failure modes from failed trajectories and retrieves tasks with similar failure patterns, thereby reshaping the training data distribution for targeted practice. Experiments on multiple benchmarks show that Role-Agent consistently improves performance, yielding an average gain of over 4% over strong baselines.

In this paper, we introduce the Soft Lamprey-Inspired Dual Environment Robot (SLIDER) and a proper modeling and optimization procedure employed to design the robot. We represent the primary fluid environment actions - inertial effects, vortex forces, and viscous dissipation - using Lighthill's theory for large-amplitude elongated bodies. For structural design parameters such as internal pressure, tail size, and body stiffness, a fast, geometrically and materially nonlinear model is developed and validated. The fluid-structure interaction equations are solved implicitly with an efficient second-order box method. A pneumatic manifold robotic system is employed to actuate SLIDER in a quiescent water tank environment, allowing cross-comparison of computational and experimental results. We find that low-frequency swimming is dominated by resistant environmental forces, whereas higher-frequency swimming is primarily affected by inertial fluid forces. Using our efficient model alongside a genetic algorithm, we co-optimize a swimming control pattern and caudal fin design (subject to SLIDER's climbing morphology) to achieve a tethered swimming speed of 21.7 +/- 0.4 cm/s (0.59 Bl/s). Furthermore, we investigate the optimization procedure for a multimodal robot performing both swimming and climbing tasks.

Planning under uncertainty requires agents to balance goal achievement with information gathering. Active inference addresses this through the Expected Free Energy (EFE), a cost function that unifies instrumental and epistemic objectives. However, existing EFE-based methods typically employ specialized optimization procedures that are difficult to extend or analyze. In this paper, we show that EFE-based planning can be formulated as Variational Free Energy minimization on a generative model augmented with epistemic priors. Our main result demonstrates that minimizing a Variational Free Energy functional with appropriately chosen priors yields a decomposition into expected plan costs (the EFE) plus a complexity term. This formulation reinforces theoretical consistency with the Free Energy Principle by casting planning as the same inferential process that governs perception and learning. We validate our approach on three environments of increasing complexity: a deterministic T-maze, a stochastic Reactivity Maze, and a partially observable MiniGrid DoorKey-8x8 environment. The experiments demonstrate that the epistemic priors induce information-seeking behavior, that the variational formulation yields policy-based inference outperforming plan-based methods under stochastic transitions, and that temporal factorization enables scalability to environments where existing tabular active inference methods cannot operate.

Large language models are increasingly used to power personal agents for everyday applications, but evaluating these agents remains a challenge. Existing benchmarks still rely on sandboxed artifacts, static task design, and coarse scoring, which hinder scalability and limit progress toward reliable personal-agent evaluation. This paper introduces STAGE-Claw, an automated framework for building and evaluating realistic personal-agent scenarios in state-based personal-computing environments. Given a task hint, STAGE-Claw automatically creates and validates a realistic benchmark task with its environment, task prompts, ground truth, and related verification programs. Agents are then evaluated in realistic operating environments, where performance is measured by the correctness of the final system state rather than only the textual response. Using STAGE-Claw, this paper creates a benchmark with 40 challenging real scenario agent tasks, evaluates 11 frontier models, and analyzes their task scores, costs, tool-call reliability, and common failure patterns. Overall, STAGE-Claw offers a scalable, state-based way to evaluate agents in realistic user scenarios.