Large Language Model (LLM)-based agentic systems, often comprising multiple models, complex tool invocations, and orchestration protocols, substantially outperform monolithic agents. Yet this very sophistication amplifies their fragility, making them more prone to system failure. Pinpointing the specific agent or step responsible for an error within long execution traces defines the task of agentic system failure attribution. Current state-of-the-art reasoning LLMs, however, remain strikingly inadequate for this challenge, with accuracy generally below 10%. To address this gap, we propose AgenTracer, the first automated framework for annotating failed multi-agent trajectories via counterfactual replay and programmed fault injection, producing the curated dataset TracerTraj. Leveraging this resource, we develop AgenTracer-8B, a lightweight failure tracer trained with multi-granular reinforcement learning, capable of efficiently diagnosing errors in verbose multi-agent interactions. On the Who&When benchmark, AgenTracer-8B outperforms giant proprietary LLMs like Gemini-2.5-Pro and Claude-4-Sonnet by up to 18.18%, setting a new standard in LLM agentic failure attribution. More importantly, AgenTracer-8B delivers actionable feedback to off-the-shelf multi-agent systems like MetaGPT and MaAS with 4.8-14.2% performance gains, empowering self-correcting and self-evolving agentic AI.

Theory of Mind (ToM) -- the ability to understand that others can have differing knowledge and goals -- enables agents to reason about others' beliefs while planning their own actions. We present a novel approach to multi-agent cooperation by implementing ToM within active inference. Unlike previous active inference approaches to multi-agent cooperation, our method neither relies on task-specific shared generative models nor requires explicit communication. In our framework, ToM-equipped agents maintain distinct representations of their own and others' beliefs and goals. ToM agents then use an extended and adapted version of the sophisticated inference tree-based planning algorithm to systematically explore joint policy spaces through recursive reasoning. We evaluate our approach through collision avoidance and foraging simulations. Results suggest that ToM agents cooperate better compared to non-ToM counterparts by being able to avoid collisions and reduce redundant efforts. Crucially, ToM agents accomplish this by inferring others' beliefs solely from observable behaviour and considering them when planning their own actions. Our approach shows potential for generalisable and scalable multi-agent systems while providing computational insights into ToM mechanisms.

Online 3D Bin Packing Problem (3D-BPP) has widespread applications in industrial automation. Existing methods usually solve the problem with limited resolution of spatial discretization, and/or cannot deal with complex practical constraints well. We propose to enhance the practical applicability of online 3D-BPP via learning on a novel hierarchical representation, packing configuration tree (PCT). PCT is a full-fledged description of the state and action space of bin packing which can support packing policy learning based on deep reinforcement learning (DRL). The size of the packing action space is proportional to the number of leaf nodes, making the DRL model easy to train and well-performing even with continuous solution space. We further discover the potential of PCT as tree-based planners in deliberately solving packing problems of industrial significance, including large-scale packing and different variations of BPP setting. A recursive packing method is proposed to decompose large-scale packing into smaller sub-trees while a spatial ensemble mechanism integrates local solutions into global. For different BPP variations with additional decision variables, such as lookahead, buffering, and offline packing, we propose a unified planning framework enabling out-of-the-box problem solving. Extensive evaluations demonstrate that our method outperforms existing online BPP baselines and is versatile in incorporating various practical constraints. The planning process excels across large-scale problems and diverse problem variations. We develop a real-world packing robot for industrial warehousing, with careful designs accounting for constrained placement and transportation stability. Our packing robot operates reliably and efficiently on unprotected pallets at 10 seconds per box. It achieves averagely 19 boxes per pallet with 57.4% space utilization for relatively large-size boxes.

This paper proposes a fully data-driven motion-planning framework for homogeneous linear multi-agent systems that operate in shared, obstacle-filled workspaces without access to explicit system models. Each agent independently learns its closed-loop behavior from experimental data by solving convex semidefinite programs that generate locally invariant ellipsoids and corresponding state-feedback gains. These ellipsoids, centered along grid-based waypoints, certify the dynamic feasibility of short-range transitions and define safe regions of operation. A sampling-based planner constructs a tree of such waypoints, where transitions are allowed only when adjacent ellipsoids overlap, ensuring invariant-to-invariant transitions and continuous safety. All agents expand their trees simultaneously and are coordinated through a space-time reservation table that guarantees inter-agent safety by preventing simultaneous occupancy and head-on collisions. Each successful edge in the tree is equipped with its own local controller, enabling execution without re-solving optimization problems at runtime. The resulting trajectories are not only dynamically feasible but also provably safe with respect to both environmental constraints and inter-agent collisions. Simulation results demonstrate the effectiveness of the approach in synthesizing synchronized, safe trajectories for multiple agents under shared dynamics and constraints, using only data and convex optimization tools.

We introduce MBTI-in-Thoughts, a framework for enhancing the effectiveness of Large Language Model (LLM) agents through psychologically grounded personality conditioning. Drawing on the Myers-Briggs Type Indicator (MBTI), our method primes agents with distinct personality archetypes via prompt engineering, enabling control over behavior along two foundational axes of human psychology, cognition and affect. We show that such personality priming yields consistent, interpretable behavioral biases across diverse tasks: emotionally expressive agents excel in narrative generation, while analytically primed agents adopt more stable strategies in game-theoretic settings. Our framework supports experimenting with structured multi-agent communication protocols and reveals that self-reflection prior to interaction improves cooperation and reasoning quality. To ensure trait persistence, we integrate the official 16Personalities test for automated verification. While our focus is on MBTI, we show that our approach generalizes seamlessly to other psychological frameworks such as Big Five, HEXACO, or En-neagram. By bridging psychological theory and LLM behavior design, we establish a foundation for psychologically enhanced AI agents without any fine-tuning.

In the game of poker, being unpredictable, or bluffing, is an essential skill. When humans play poker, they bluff. However, most works on computer-poker focus on performance metrics such as win rates, while bluffing is overlooked. In this paper we study whether two popular algorithms, DQN (based on reinforcement learning) and CFR (based on game theory), exhibit bluffing behavior in Leduc Hold'em, a simplified version of poker. We designed an experiment where we let the DQN and CFR agent play against each other while we log their actions. We find that both DQN and CFR exhibit bluffing behavior, but they do so in different ways. Although both attempt to perform bluffs at different rates, the percentage of successful bluffs (where the opponent folds) is roughly the same. This suggests that bluffing is an essential aspect of the game, not of the algorithm. Future work should look at different bluffing styles and at the full game of poker.

Image provided by the authors – generated using Gemini. For many of us, artificial intelligence (AI) has become part of everyday life, and the rate at which we assign previously human roles to AI systems shows no signs of slowing down. AI systems are the crucial ingredients of many technologies -- e.g., self-driving cars, smart urban planning, digital assistants -- across a growing number of domains. At the core of many of these technologies are autonomous agents -- systems designed to act on behalf of humans and make decisions without direct supervision. In order to act effectively in the real world, these agents must be capable of carrying out a wide range of tasks despite possibly unpredictable environmental conditions, which often requires some form of machine learning (ML) for achieving adaptive behaviour.

As AI agents become increasingly autonomous and capable, ensuring their security against vulnerabilities such as prompt injection becomes critical. This paper explores the use of information-flow control (IFC) to provide security guarantees for AI agents. We present a formal model to reason about the security and expressiveness of agent planners. Using this model, we characterize the class of properties enforceable by dynamic taint-tracking and construct a taxonomy of tasks to evaluate security and utility trade-offs of planner designs. Informed by this exploration, we present Fides, a planner that tracks confidentiality and integrity labels, deterministically enforces security policies, and introduces novel primitives for selectively hiding information. Its evaluation in AgentDojo demonstrates that this approach enables us to complete a broad range of tasks with security guarantees. A tutorial to walk readers through the the concepts introduced in the paper can be found at https://github.com/microsoft/fides

Agentic workflows commonly coordinate multiple models and tools with complex control logic. They are quickly becoming the dominant paradigm for AI applications. However, serving them remains inefficient with today's frameworks. The key problem is that they expose workflows as opaque sequences of model and tool calls that tightly couple agent logic with model and hardware choices. Often, these workflow components are fragmented across different entities, preventing systems from reasoning about trade-offs across accuracy, latency, energy, and cost. This leads to resource waste and degraded service-level objectives (SLOs). We present Murakkab, a resource-efficient serving system for agentic workflows. Murakkab introduces a declarative abstraction that decouples workflow specification from execution configuration. A profile-guided optimizer and adaptive runtime jointly manage the full stack: orchestrating workflow components, mapping them to models and hardware, and dynamically reconfiguring execution to satisfy user-defined SLOs. By exposing the internal structure of agentic workflows, Murakkab enables cross-layer optimization that existing frameworks and cloud schedulers cannot achieve. Our evaluation on diverse workflows shows that Murakkab reduces GPU usage by up to 2.8$\times$, energy consumption by 3.7$\times$, and cost by 4.3$\times$ while maintaining SLOs.

--Resource allocation is crucial for the performance optimization of cloud-assisted multi-agent intelligence. Traditional methods often overlook agents' diverse computational capabilities and complex operating environments, leading to inefficient and unfair resource distribution. T o address this, we open-sourced Fair-Synergy, an algorithmic framework that utilizes the concave relationship between the agents' accuracy and the system resources to ensure fair resource allocation across fleet intelligence. We extend traditional allocation approaches to encompass a multidimensional machine learning utility landscape defined by model parameters, training data volume, and task complexity. We evaluate Fair-Synergy with advanced vision and language models such as BERT, VGG16, MobileNet, and ResNets on datasets including MNIST, CIF AR-10, CIF AR-100, BDD, and GLUE. We demonstrate that Fair-Synergy outperforms standard benchmarks by up to 25% in multi-agent inference and 11% in multi-agent learning settings. Also, we explore how the level of fairness affects the least advantaged, most advantaged, and average agents, providing insights for equitable fleet intelligence.