This paper proposes an Interactive Inference Behavior Tree (IIBT) framework that integrates behavior trees (BTs) with active inference under the free energy principle for distributed multi-robot decision-making. The proposed IIBT node extends conventional BTs with probabilistic reasoning, enabling online joint planning and execution across multiple robots. It remains fully compatible with standard BT architectures, allowing seamless integration into existing multi-robot control systems. Within this framework, multi-robot cooperation is formulated as a free-energy minimization process, where each robot dynamically updates its preference matrix based on perceptual inputs and peer intentions, thereby achieving adaptive coordination in partially observable and dynamic environments. The proposed approach is validated through both simulation and real-world experiments, including a multi-robot maze navigation and a collaborative manipulation task, compared against traditional BTs(https://youtu.be/KX_oT3IDTf4). Experimental results demonstrate that the IIBT framework reduces BT node complexity by over 70%, while maintaining robust, interpretable, and adaptive cooperative behavior under environmental uncertainty.

As agentic AI systems increasingly operate autonomously, establishing trust through verifiable evaluation becomes critical. Yet existing benchmarks lack the transparency and auditability needed to assess whether agents behave reliably. We present DrawingBench, a verification framework for evaluating the trustworthiness of agentic LLMs through spatial reasoning tasks that require generating sequences of low-level GUI actions. Unlike opaque evaluations, DrawingBench provides transparent, rule-based assessment: 8 objective criteria enable reproducible scoring, while action-level inspection allows stakeholders to audit agent behavior. Our framework comprises 250 diverse prompts across 20 categories and 4 difficulty levels, deterministic evaluation metrics, and an external oversight mechanism through multi-turn feedback that enables human control over agent refinement. Evaluating four state-of-the-art LLMs (Claude-4 Sonnet, GPT-4.1, GPT-4.1-mini, Gemini-2.5 Flash) across 1,000 tests, we establish both capabilities and limitations: models achieved 92.8% perfect performance with structured external feedback driving significant improvements (average +3.2%, up to +32.8% for complex scenes), but systematic error patterns emerged in tool state management and long-horizon planning. Notably, specification clarity proved more important than task complexity -- models achieved 100% perfect performance when given explicit, verifiable criteria. These findings demonstrate that transparent evaluation frameworks can establish trust in agentic systems, with external oversight proving more reliable than self-correction for guiding agent behavior. Our open-source framework provides a template for trustworthy agent assessment. Code and data: https://github.com/hyunjun1121/DrawingBench

Autonomous agents that accomplish complex computer tasks with minimal human interventions have the potential to transform human-computer interaction, significantly enhancing accessibility and productivity. However, existing benchmarks either lack an interactive environment or are limited to environments specific to certain applications or domains, failing to reflect the diverse and complex nature of real-world computer use, thereby limiting the scope of tasks and agent scalability. To address this issue, we introduce OSWorld, the first-of-its-kind scalable, real computer environment for multimodal agents, supporting task setup, execution-based evaluation, and interactive learning across various operating systems such as Ubuntu, Windows, and macOS. OSWorld can serve as a unified, integrated computer environment for assessing open-ended computer tasks that involve arbitrary applications. Building upon OSWorld, we create a benchmark of 369 computer tasks involving real web and desktop apps in open domains, OS file I/O, and workflows spanning multiple applications. Each task example is derived from real-world computer use cases and includes a detailed initial state setup configuration and a custom execution-based evaluation script for reliable, reproducible evaluation. Extensive evaluation of state-of-the-art LLM/VLM-based agents on OSWorld reveals significant deficiencies in their ability to serve as computer assistants. While humans can accomplish over 72.36% of the tasks, the best model achieves only 12.24% success, primarily struggling with GUI grounding and operational knowledge. Comprehensive analysis using OSWorld provides valuable insights for developing multimodal generalist agents that were not possible with previous benchmarks. Our code, environment, baseline models, and data are publicly available at https://os-world.github.io.

Reinforcement learning often requires millions of examples to produce static, black-box models. In contrast, interactive task learning (ITL) emphasizes incremental knowledge acquisition from limited instruction provided by humans in modalities such as natural language. However, in practice, ITL systems often suffers from brittle, error-prone language parsing. Large language models (LLMs) are resistant to brittleness but are not interpretable and cannot learn incrementally. We present VAL, an ITL system with a new philosophy for LLM/symbolic integration. By using LLMs only for specific tasks -- such as predicate and argument selection -- within an algorithmic framework, VAL reaps the benefits of LLMs to support interactive learning of hierarchical task knowledge from natural language. Acquired knowledge is human interpretable and generalizes to support execution of novel tasks without additional training. We studied users' interactions with VAL in a video game setting, finding that most users could successfully teach VAL using language they felt was natural.

TMs are a pattern recognition approach that uses finite state machines for learning and propositional logic to represent patterns. In addition to being natively interpretable, they have provided competitive accuracy for various tasks. In this paper, we increase the computing power of TMs by proposing a first-order logic-based framework with Herbrand semantics. The resulting TM is relational and can take advantage of logical structures appearing in natural language, to learn rules that represent how actions and consequences are related in the real world. The outcome is a logic program of Horn clauses, bringing in a structured view of unstructured data. In closed-domain question-answering, the first-order representation produces 10x more compact KBs, along with an increase in answering accuracy from 94.83% to 99.48%. The approach is further robust towards erroneous, missing, and superfluous information, distilling the aspects of a text that are important for real-world understanding.

When designing plans in engineering, it is often necessary to consider attributes associated to objects, e.g. the location of a robot. Our aim in this paper is to incorporate attributes into existing categorical formalisms for planning, namely those based on symmetric monoidal categories and string diagrams. To accomplish this, we define a notion of a "symmetric monoidal category with attributes." This is a symmetric monoidal category in which objects are equipped with retrievable information and where the interactions between objects and information are governed by an "attribute structure." We discuss examples and semantics of such categories in the context of robotics to illustrate our definition.

We extend probabilistic action language pBC+ with the notion of utility as in decision theory. The semantics of the extended pBC+ can be defined as a shorthand notation for a decision-theoretic extension of the probabilistic answer set programming language LPMLN. Alternatively, the semantics of pBC+ can also be defined in terms of Markov Decision Process (MDP), which in turn allows for representing MDP in a succinct and elaboration tolerant way as well as to leverage an MDP solver to compute pBC+. The idea led to the design of the system pbcplus2mdp, which can find an optimal policy of a pBC+ action description using an MDP solver.

In this paper, we show how a planning algorithm can be used to automatically create and update a Behavior Tree (BT), controlling a robot in a dynamic environment. The planning part of the algorithm is based on the idea of back chaining. Starting from a goal condition we iteratively select actions to achieve that goal, and if those actions have unmet preconditions, they are extended with actions to achieve them in the same way. The fact that BTs are inherently modular and reactive makes the proposed solution blend acting and planning in a way that enables the robot to efficiently react to external disturbances. If an external agent undoes an action the robot reexecutes it without re-planning, and if an external agent helps the robot, it skips the corresponding actions, again without replanning. We illustrate our approach in two different robotics scenarios.

Interactions within human societies are usually regulated by social norms. If robots are to be accepted into human society, it is essential that they are aware of and capable of reasoning about social norms. In this paper, we focus on how to represent social norms in societies with humans and robots, and how artificial agents such as robots can reason about social norms in order to plan appropriate behavior. We use the notion of institution as a way to formally define and encapsulate norms. We provide a formal framework built around the notion of institution. The framework distinguishes between abstract norms and their semantics in a concrete domain, hence allowing the use of the same institution across physical domains and agent types. It also provides a formal computational framework for norm verification, planning, and plan execution in a domain.

This paper is directed towards combining Pearl's structural-model approach to causal reasoning with high-level formalisms for reasoning about actions. More precisely, we present a combination of Pearl's structural-model approach with Poole's independent choice logic. We show how probabilistic theories in the independent choice logic can be mapped to probabilistic causal models. This mapping provides the independent choice logic with appealing concepts of causality and explanation from the structural-model approach. We illustrate this along Halpern and Pearl's sophisticated notions of actual cause, explanation, and partial explanation. This mapping also adds first-order modeling capabilities and explicit actions to the structural-model approach.