Proper initialization is crucial for any system, particularly in multi-agent systems (MAS), where it plays a pivotal role in determining both the system's efficiency and effectiveness. However, existing MAS initialization methods do not fully account for the collaborative needs of the generated agents in subsequent stages. Inspired by the principles of effective team composition, we propose AgentInit, which aims to optimize the structure of agent teams. Specifically, in addition to multi-round interactions and reflections between agents during agent generation, AgentInit incorporates a Natural Language to Format mechanism to ensure consistency and standardization. Balanced team selection strategies using Pareto principles are subsequently applied to jointly consider agent team diversity and task relevance to promote effective and efficient collaboration and enhance overall system performance. Experiments show that AgentInit consistently outperforms state-of-the-art initialization methods and pre-defined strategies across various frameworks and tasks, achieving an overall performance improvement of up to 1.2 and 1.6, respectively, while also significantly reducing token consumption. Further analysis confirms its strong transferability to similar tasks and verifies the effectiveness of its key components, demonstrating its capability and adaptability as a reliable MAS initialization method. Source code and models are available at https://github.com/1737423697/AgentInit.
Large language models (LLMs) are increasingly trained with reinforcement learning (RL) as autonomous agents that reason and act over long horizons in interactive environments. However, sparse and sometimes unverifiable rewards make temporal credit assignment extremely challenging. Recent work attempts to integrate process supervision into agent learning but suffers from biased annotation, reward hacking, high-variance from overly fine-grained signals or failtures when state overlap is rare. We therefore introduce Online Process Reward Learning (OPRL), a general credit-assignment strategy for agentic RL that integrates seamlessly with standard on-policy algorithms without relying on additional rollouts or explicit step labels. In OPRL, we optimize an implicit process reward model (PRM) alternately with the agent's policy to transform trajectory preferences into implicit step rewards through a trajectory-based DPO objective. These step rewards are then used to compute step-level advantages, which are combined with episode-level advantages from outcome rewards for policy update, creating a self-reinforcing loop. Theoretical findings guarantee that the learned step rewards are consistent with trajectory preferences and act as potential-based shaping rewards, providing bounded gradients to stabilize training. Empirically, we evaluate OPRL on three distinct agent benmarks, including WebShop and VisualSokoban, as well as open-ended social interactions with unverfiable rewards in SOTOPIA. Crucially, OPRL shows superior performance over frontier LLMs and strong RL baselines across domains, achieving state-of-the-art results with higher sample-efficiency and lower variance during training. Further analysis also demonstrates the efficient exploration by OPRL using fewer actions, underscoring its potential for agentic learning in real-world scenarios.
The use of natural language (NL) test cases for validating graphical user interface (GUI) applications is emerging as a promising direction to manually written executable test scripts, which are costly to develop and difficult to maintain. Recent advances in large language models (LLMs) have opened the possibility of the direct execution of NL test cases by LLM agents. This paper investigates this direction, focusing on the impact on NL test case unsoundness and on test case execution consistency. NL test cases are inherently unsound, as they may yield false failures due to ambiguous instructions or unpredictable agent behaviour. Furthermore, repeated executions of the same NL test case may lead to inconsistent outcomes, undermining test reliability. To address these challenges, we propose an algorithm for executing NL test cases with guardrail mechanisms and specialised agents that dynamically verify the correct execution of each test step. We introduce measures to evaluate the capabilities of LLMs in test execution and one measure to quantify execution consistency. We propose a definition of weak unsoundness to characterise contexts in which NL test case execution remains acceptable, with respect to the industrial quality levels Six Sigma. Our experimental evaluation with eight publicly available LLMs, ranging from 3B to 70B parameters, demonstrates both the potential and current limitations of current LLM agents for GUI testing. Our experiments show that Meta Llama 3.1 70B demonstrates acceptable capabilities in NL test case execution with high execution consistency (above the level 3-sigma). We provide prototype tools, test suites, and results.
Given the widespread use of deep learning models in safety-critical applications, ensuring that the decisions of such models are robust against adversarial exploitation is of fundamental importance. In this thesis, we discuss recent progress toward designing algorithms that exhibit desirable robustness properties. First, we discuss the problem of adversarial examples in computer vision, for which we introduce new technical results, training paradigms, and certification algorithms. Next, we consider the problem of domain generalization, wherein the task is to train neural networks to generalize from a family of training distributions to unseen test distributions. We present new algorithms that achieve state-of-the-art generalization in medical imaging, molecular identification, and image classification. Finally, we study the setting of jailbreaking large language models (LLMs), wherein an adversarial user attempts to design prompts that elicit objectionable content from an LLM. We propose new attacks and defenses, which represent the frontier of progress toward designing robust language-based agents.
Large Language Models (LLMs) have been widely adopted as task planners for AI agents in sequential decision-making problems, leveraging their extensive world knowledge. However, the gap between their general knowledge and environment-specific requirements often leads to inaccurate plans. To address this, existing approaches rely on frequent LLM queries to iteratively refine plans based on immediate environmental feedback, which incurs substantial query costs. However, this refinement is typically guided by short-term environmental feedback, limiting LLMs from developing plans aligned with long-term rewards. We propose Code Driven Planning with Domain-Adaptive Critic (CoPiC). Instead of relying on frequent queries, CoPiC employs LLMs to generate a diverse set of high-level planning programs, which iteratively produce and refine candidate plans. A trained domain-adaptive critic then evaluates these candidates and selects the one most aligned with long-term rewards for execution. Using high-level planning programs as planner and domain-adaptive critic as estimator, CoPiC improves planning while significantly reducing query costs. Results in ALFWorld, NetHack, and StarCraft II Unit Building show that CoPiC outperforms advanced LLM-based baselines, AdaPlanner and Reflexion, achieving an average (1) 23.33% improvement in success rate and (2) 91.27% reduction in query costs.
Online social networks offer a valuable lens to analyze both individual and collective phenomena. Researchers often use simulators to explore controlled scenarios, and the integration of Large Language Models (LLMs) makes these simulations more realistic by enabling agents to understand and generate natural language content. In this work, we investigate the behavior of LLM-based agents in a simulated microblogging social network. We initialize agents with realistic profiles calibrated on real-world online conversations from the 2022 Italian political election and extend an existing simulator by introducing mechanisms for opinion modeling. We examine how LLM agents simulate online conversations, interact with others, and evolve their opinions under different scenarios. Our results show that LLM agents generate coherent content, form connections, and build a realistic social network structure. However, their generated content displays less heterogeneity in tone and toxicity compared to real data. We also find that LLM-based opinion dynamics evolve over time in ways similar to traditional mathematical models. Varying parameter configurations produces no significant changes, indicating that simulations require more careful cognitive modeling at initialization to replicate human behavior more faithfully. Overall, we demonstrate the potential of LLMs for simulating user behavior in social environments, while also identifying key challenges in capturing heterogeneity and complex dynamics.
Driven by the rapid advancements of Large Language Models (LLMs), LLM-based agents have emerged as powerful intelligent systems capable of human-like cognition, reasoning, and interaction. These agents are increasingly being deployed across diverse real-world applications, including student education, scientific research, and financial analysis. However, despite their remarkable potential, LLM-based agents remain vulnerable to hallucination issues, which can result in erroneous task execution and undermine the reliability of the overall system design. Addressing this critical challenge requires a deep understanding and a systematic consolidation of recent advances on LLM-based agents. To this end, we present the first comprehensive survey of hallucinations in LLM-based agents. By carefully analyzing the complete workflow of agents, we propose a new taxonomy that identifies different types of agent hallucinations occurring at different stages. Furthermore, we conduct an in-depth examination of eighteen triggering causes underlying the emergence of agent hallucinations. Through a detailed review of a large number of existing studies, we summarize approaches for hallucination mitigation and detection, and highlight promising directions for future research. We hope this survey will inspire further efforts toward addressing hallucinations in LLM-based agents, ultimately contributing to the development of more robust and reliable agent systems.
Tool-augmented large language models (LLMs) are usually trained with supervised imitation or coarse-grained reinforcement learning that optimizes single tool calls. Current self-reflection practices rely on heuristic prompts or one-way reasoning: the model is urged to 'think more' instead of learning error diagnosis and repair. This is fragile in multi-turn interactions; after a failure the model often repeats the same mistake. We propose structured reflection, which turns the path from error to repair into an explicit, controllable, and trainable action. The agent produces a short yet precise reflection: it diagnoses the failure using evidence from the previous step and then proposes a correct, executable follow-up call. For training we combine DAPO and GSPO objectives with a reward scheme tailored to tool use, optimizing the stepwise strategy Reflect, then Call, then Final. To evaluate, we introduce Tool-Reflection-Bench, a lightweight benchmark that programmatically checks structural validity, executability, parameter correctness, and result consistency. Tasks are built as mini trajectories of erroneous call, reflection, and corrected call, with disjoint train and test splits. Experiments on BFCL v3 and Tool-Reflection-Bench show large gains in multi-turn tool-call success and error recovery, and a reduction of redundant calls. These results indicate that making reflection explicit and optimizing it directly improves the reliability of tool interaction and offers a reproducible path for agents to learn from failure.
Keyphrase extraction is a fundamental task in natural language processing. However, existing unsupervised prompt-based methods for Large Language Models (LLMs) often rely on single-stage inference pipelines with uniform prompting, regardless of document length or LLM backbone. Such one-size-fits-all designs hinder the full exploitation of LLMs' reasoning and generation capabilities, especially given the complexity of keyphrase extraction across diverse scenarios. To address these challenges, we propose MAPEX, the first framework that introduces multi-agent collaboration into keyphrase extraction. MAPEX coordinates LLM-based agents through modules for expert recruitment, candidate extraction, topic guidance, knowledge augmentation, and post-processing. A dual-path strategy dynamically adapts to document length: knowledge-driven extraction for short texts and topic-guided extraction for long texts. Extensive experiments on six benchmark datasets across three different LLMs demonstrate its strong generalization and universality, outperforming the state-of-the-art unsupervised method by 2.44\% and standard LLM baselines by 4.01\% in F1@5 on average. Code is available at https://github.com/NKU-LITI/MAPEX.
Large language models (LLMs) have achieved remarkable progress in code generation. However, existing benchmarks mainly formalize the task as a static, single-turn problem, overlooking the stepwise requirement changes and iterative workflows in real-world software development. This mismatch limits the understanding of how well LLMs can support real-world development workflows. Constructing such iterative benchmarks is challenging due to the lack of public interaction traces and the difficulty of creating discriminative, turn-specific test cases. To bridge this gap, we present SR-Eval, a benchmark specifically designed to assess LLMs on iterative code generation under Stepwise requirements Refinement. SR-Eval spans both function-level and repository-level tasks in Python and Java, enabling fine-grained and progressive evaluation across evolving requirements. The construction of SR-Eval follows a carefully designed pipeline that first leverages a multi-agent-based requirement generation method to simulate the development process and recover the multi-round interaction process from final requirements, then employs a semantic-aware discriminative test case generation component to ensure discriminative and consistent evaluation at each turn. SR-Eval comprises 443 multi-turn tasks and 1,857 questions at both function and repository levels. Using SR-Eval, we evaluate 11 representative LLMs with three prompting strategies that simulate different usage patterns. Results show that iterative code generation under stepwise requirement refinement remains highly challenging: the best-performing model achieves only 22.67% completion rate on function-level tasks and 20.00% on repository-level tasks. We further observe that prompting strategies substantially influence performance, highlighting the need for the development of advanced methods.
The Agent Directory Service (ADS) is a distributed directory for the discovery of AI agent capabilities, metadata, and provenance. It leverages content-addressed storage, hierarchical taxonomies, and cryptographic signing to enable efficient, verifiable, and multi-dimensional discovery across heterogeneous Multi-Agent Systems (MAS). Built on the Open Agentic Schema Framework (OASF), ADS decouples capability indexing from content location through a two-level mapping realized over a Kademlia-based Distributed Hash Table (DHT). It reuses mature OCI / ORAS infrastructure for artifact distribution, integrates Sigstore for provenance, and supports schema-driven extensibility for emerging agent modalities (LLM prompt agents, MCP servers, A2A-enabled components). This paper formalizes the architectural model, describes storage and discovery layers, explains security and performance properties, and positions ADS within the broader landscape of emerging agent registry and interoperability initiatives.
Large Language Model-based agents(LLM-based agents) are increasingly deployed in customer service, yet they often forget across sessions, repeat errors, and lack mechanisms for continual self-improvement. This makes them unreliable in dynamic settings where stability and consistency are critical. To better evaluate these properties, we emphasize two indicators: task success rate as a measure of overall effectiveness, and consistency metrics such as Pass$^k$ to capture reliability across multiple trials. To address the limitations of existing approaches, we propose MemOrb, a lightweight and plug-and-play verbal reinforcement memory layer that distills multi-turn interactions into compact strategy reflections. These reflections are stored in a shared memory bank and retrieved to guide decision-making, without requiring any fine-tuning. Experiments show that MemOrb significantly improves both success rate and stability, achieving up to a 63 percentage-point gain in multi-turn success rate and delivering more consistent performance across repeated trials. Our results demonstrate that structured reflection is a powerful mechanism for enhancing long-term reliability of frozen LLM agents in customer service scenarios.
As data continues to grow in scale and complexity, preparing, transforming, and analyzing it remains labor-intensive, repetitive, and difficult to scale. Since data contains knowledge and AI learns knowledge from it, the alignment between AI and data is essential. However, data is often not structured in ways that are optimal for AI utilization. Moreover, an important question arises: how much knowledge can we pack into data through intensive data operations? Autonomous data agents (DataAgents), which integrate LLM reasoning with task decomposition, action reasoning and grounding, and tool calling, can autonomously interpret data task descriptions, decompose tasks into subtasks, reason over actions, ground actions into python code or tool calling, and execute operations. Unlike traditional data management and engineering tools, DataAgents dynamically plan workflows, call powerful tools, and adapt to diverse data tasks at scale. This report argues that DataAgents represent a paradigm shift toward autonomous data-to-knowledge systems. DataAgents are capable of handling collection, integration, preprocessing, selection, transformation, reweighing, augmentation, reprogramming, repairs, and retrieval. Through these capabilities, DataAgents transform complex and unstructured data into coherent and actionable knowledge. We first examine why the convergence of agentic AI and data-to-knowledge systems has emerged as a critical trend. We then define the concept of DataAgents and discuss their architectural design, training strategies, as well as the new skills and capabilities they enable. Finally, we call for concerted efforts to advance action workflow optimization, establish open datasets and benchmark ecosystems, safeguard privacy, balance efficiency with scalability, and develop trustworthy DataAgent guardrails to prevent malicious actions.
The exponential growth of scientific literature poses unprecedented challenges for researchers attempting to synthesize knowledge across rapidly evolving fields. We present \textbf{Agentic AutoSurvey}, a multi-agent framework for automated survey generation that addresses fundamental limitations in existing approaches. Our system employs four specialized agents (Paper Search Specialist, Topic Mining \& Clustering, Academic Survey Writer, and Quality Evaluator) working in concert to generate comprehensive literature surveys with superior synthesis quality. Through experiments on six representative LLM research topics from COLM 2024 categories, we demonstrate that our multi-agent approach achieves significant improvements over existing baselines, scoring 8.18/10 compared to AutoSurvey's 4.77/10. The multi-agent architecture processes 75--443 papers per topic (847 total across six topics) while targeting high citation coverage (often $\geq$80\% on 75--100-paper sets; lower on very large sets such as RLHF) through specialized agent orchestration. Our 12-dimension evaluation captures organization, synthesis integration, and critical analysis beyond basic metrics. These findings demonstrate that multi-agent architectures represent a meaningful advancement for automated literature survey generation in rapidly evolving scientific domains.
To assist users in complex tasks, LLMs generate plans: step-by-step instructions towards a goal. While alignment methods aim to ensure LLM plans are helpful, they train (RLHF) or evaluate (ChatbotArena) on what users prefer, assuming this reflects what helps them. We test this with Planorama: an interface where 126 users answer 300 multi-step questions with LLM plans. We get 4388 plan executions and 5584 comparisons to measure plan helpfulness (QA success) and user preferences on plans, and recreate the setup in agents and reward models to see if they simulate or prefer what helps users. We expose: 1) user/model preferences and agent success do not accurately predict which plans help users, so common alignment feedback can misalign with helpfulness; 2) this gap is not due to user-specific preferences, as users are similarly successful when using plans they prefer/disprefer; 3) surface-level cues like brevity and question similarity strongly link to preferences, but such biases fail to predict helpfulness. In all, we argue aligning helpful LLMs needs feedback from real user interactions, not just preferences of what looks helpful, so we discuss the plan NLP researchers can execute to solve this problem.
Large language models (LLMs)-based code generation for robotic manipulation has recently shown promise by directly translating human instructions into executable code, but existing methods remain noisy, constrained by fixed primitives and limited context windows, and struggle with long-horizon tasks. While closed-loop feedback has been explored, corrected knowledge is often stored in improper formats, restricting generalization and causing catastrophic forgetting, which highlights the need for learning reusable skills. Moreover, approaches that rely solely on LLM guidance frequently fail in extremely long-horizon scenarios due to LLMs' limited reasoning capability in the robotic domain, where such issues are often straightforward for humans to identify. To address these challenges, we propose a human-in-the-loop framework that encodes corrections into reusable skills, supported by external memory and Retrieval-Augmented Generation with a hint mechanism for dynamic reuse. Experiments on Ravens, Franka Kitchen, and MetaWorld, as well as real-world settings, show that our framework achieves a 0.93 success rate (up to 27% higher than baselines) and a 42% efficiency improvement in correction rounds. It can robustly solve extremely long-horizon tasks such as "build a house", which requires planning over 20 primitives.
Multimodal semantic learning plays a critical role in embodied intelligence, especially when robots perceive their surroundings, understand human instructions, and make intelligent decisions. However, the field faces technical challenges such as effective fusion of heterogeneous data and computational efficiency in resource-constrained environments. To address these challenges, this study proposes the lightweight LCMF cascaded attention framework, introducing a multi-level cross-modal parameter sharing mechanism into the Mamba module. By integrating the advantages of Cross-Attention and Selective parameter-sharing State Space Models (SSMs), the framework achieves efficient fusion of heterogeneous modalities and semantic complementary alignment. Experimental results show that LCMF surpasses existing multimodal baselines with an accuracy of 74.29% in VQA tasks and achieves competitive mid-tier performance within the distribution cluster of Large Language Model Agents (LLM Agents) in EQA video tasks. Its lightweight design achieves a 4.35-fold reduction in FLOPs relative to the average of comparable baselines while using only 166.51M parameters (image-text) and 219M parameters (video-text), providing an efficient solution for Human-Robot Interaction (HRI) applications in resource-constrained scenarios with strong multimodal decision generalization capabilities.
Compared to traditional models, agentic AI represents a highly valuable target for potential attackers as they possess privileged access to data sources and API tools, which are traditionally not incorporated into classical agents. Unlike a typical software application residing in a Demilitarized Zone (DMZ), agentic LLMs consciously rely on nondeterministic behavior of the AI (only defining a final goal, leaving the path selection to LLM). This characteristic introduces substantial security risk to both operational security and information security. Most common existing defense mechanism rely on detection of malicious intent and preventing it from reaching the LLM agent, thus protecting against jailbreak attacks such as prompt injection. In this paper, we present an alternative approach, LLMZ+, which moves beyond traditional detection-based approaches by implementing prompt whitelisting. Through this method, only contextually appropriate and safe messages are permitted to interact with the agentic LLM. By leveraging the specificity of context, LLMZ+ guarantees that all exchanges between external users and the LLM conform to predefined use cases and operational boundaries. Our approach streamlines the security framework, enhances its long-term resilience, and reduces the resources required for sustaining LLM information security. Our empirical evaluation demonstrates that LLMZ+ provides strong resilience against the most common jailbreak prompts. At the same time, legitimate business communications are not disrupted, and authorized traffic flows seamlessly between users and the agentic LLM. We measure the effectiveness of approach using false positive and false negative rates, both of which can be reduced to 0 in our experimental setting.
Large language models (LLMs) have demonstrated strong reasoning and tool-use capabilities, yet they often fail in real-world tool-interactions due to incorrect parameterization, poor tool selection, or misinterpretation of user intent. These issues often stem from an incomplete understanding of user goals and inadequate comprehension of tool documentation. While Chain-of-Thought (CoT) prompting has proven effective for enhancing reasoning in general contexts, our analysis reveals that free-form CoT is insufficient and sometimes counterproductive for structured function-calling tasks. To address this, we introduce a curriculum-inspired framework that leverages structured reasoning templates to guide LLMs through more deliberate step-by-step instructions for generating function callings. Experimental results show that our method reduces tool-use errors, achieving 3-12% relative improvements over strong baselines across diverse model series and approaches. Moreover, our framework enhances the robustness, interpretability, and transparency of tool-using agents, advancing the development of more reliable AI assistants for real-world applications.
Large Language Models (LLMs) show strong reasoning abilities but rely on internalized knowledge that is often insufficient, outdated, or incorrect when trying to answer a question that requires specific domain knowledge. Knowledge Graphs (KGs) provide structured external knowledge, yet their complexity and multi-hop reasoning requirements make integration challenging. We present ARK-V1, a simple KG-agent that iteratively explores graphs to answer natural language queries. We evaluate several not fine-tuned state-of-the art LLMs as backbones for ARK-V1 on the CoLoTa dataset, which requires both KG-based and commonsense reasoning over long-tail entities. ARK-V1 achieves substantially higher conditional accuracies than Chain-of-Thought baselines, and larger backbone models show a clear trend toward better coverage, correctness, and stability.
We explore whether techniques from AI can help discover new combinatorial structures that improve on known limits on efficient algorithms. Specifically, we use AlphaEvolve (an LLM coding agent) to study two settings: a) Average-case hardness for MAX-CUT and MAX-Independent Set: We improve a recent result of Kunisky and Yu to obtain near-optimal upper and (conditional) lower bounds on certification algorithms for MAX-CUT and MAX-Independent Set on random 3- and 4-regular graphs. Our improved lower bounds are obtained by constructing nearly extremal Ramanujan graphs on as many as $163$ nodes, using AlphaEvolve. Additionally, via analytical arguments we strengthen the upper bounds to settle the computational hardness of these questions up to an error in the third decimal place. b) Worst-case Hardness of Approximation for MAX-k-CUT: We obtain new inapproximability results, proving that it is NP-hard to approximate MAX-4-CUT and MAX-3-CUT within factors of $0.987$ and $0.9649$ respectively, using AlphaEvolve to discover new gadget reductions. Our MAX-4-CUT result improves upon the SOTA of $0.9883$, and our MAX-3-CUT result improves on the current best gadget-based inapproximability result of $0.9853$, but falls short of improving the SOTA of $16/17$ that relies on a custom PCP, rather than a gadget reduction from "standard" H{\aa}stad-style PCPs. A key technical challenge we faced: verifying a candidate construction produced by AlphaEvolve is costly (often requiring exponential time). In both settings above, our results were enabled by using AlphaEvolve itself to evolve the verification procedure to be faster (sometimes by $10,000\times$). We conclude with a discussion of norms by which to assess the assistance from AI in developing proofs.
Large Language Models (LLMs) are increasingly used for social simulation, where populations of agents are expected to reproduce human-like collective behavior. However, we find that many recent studies adopt experimental designs that systematically undermine the validity of their claims. From a survey of over 40 papers, we identify six recurring methodological flaws: agents are often homogeneous (Profile), interactions are absent or artificially imposed (Interaction), memory is discarded (Memory), prompts tightly control outcomes (Minimal-Control), agents can infer the experimental hypothesis (Unawareness), and validation relies on simplified theoretical models rather than real-world data (Realism). For instance, GPT-4o and Qwen-3 correctly infer the underlying social experiment in 53.1% of cases when given instructions from prior work-violating the Unawareness principle. We formalize these six requirements as the PIMMUR principles and argue they are necessary conditions for credible LLM-based social simulation. To demonstrate their impact, we re-run five representative studies using a framework that enforces PIMMUR and find that the reported social phenomena frequently fail to emerge under more rigorous conditions. Our work establishes methodological standards for LLM-based multi-agent research and provides a foundation for more reliable and reproducible claims about "AI societies."
Intelligent systems have traditionally been designed as tools rather than collaborators, often lacking critical characteristics that collaboration partnerships require. Recent advances in large language model (LLM) agents open new opportunities for human-LLM-agent collaboration by enabling natural communication and various social and cognitive behaviors. Yet it remains unclear whether principles of computer-mediated collaboration established in HCI and CSCW persist, change, or fail when humans collaborate with LLM agents. To support systematic investigations of these questions, we introduce an open and configurable research platform for HCI researchers. The platform's modular design allows seamless adaptation of classic CSCW experiments and manipulation of theory-grounded interaction controls. We demonstrate the platform's effectiveness and usability through two case studies: (1) re-implementing the classic human-human-collaboration task Shape Factory as a between-subject human-agent-collaboration experiment with 16 participants, and (2) a participatory cognitive walkthrough with five HCI researchers to refine workflows and interfaces for experiment setup and analysis.
Recent advances in GUI agents have achieved remarkable grounding and action-prediction performance, yet existing models struggle with unreliable reward signals and limited online trajectory generation. In this paper, we introduce Orcust, a framework that integrates Principle-Constrained Reward Modeling (PCRM) and Online VM-Grounded Trajectory Construction (OVTC) to enhance reasoning reliability and data efficiency in interactive GUI tasks. We leverages environment-verifiable and LLM-derived principle to enforce interpretable reward signals that constrain long chain-of-thought reasoning and rule-based feedback. OVTC spins up instrumented virtual machines to autonomously collect structured GUI interaction trajectories with explicit procedural and structural objectives, enabling the training of a stepwise reward model that robustly captures human preferences and adheres to task-specific constraints. Extensive experiments on standard GUI benchmarks covering perceptual grounding, foundational operations, and end-to-end task execution reveal that Orcust achieves state-of-the-art performance, improving by 22.2\% on ScreenSpot and 23.9\% on ScreenSpot-Pro over the base model (i.e. Qwen2.5-VL-7B). The results demonstrate Orcust's effectiveness in enhancing the reasoning, adaptability and scalability of GUI agents across various environments and task complexities.
Security practitioners face growing challenges in exploit assessment, as public vulnerability repositories are increasingly populated with inconsistent and low-quality exploit artifacts. Existing scoring systems, such as CVSS and EPSS, offer limited support for this task. They either rely on theoretical metrics or produce opaque probability estimates without assessing whether usable exploit code exists. In practice, security teams often resort to manual triage of exploit repositories, which is time-consuming, error-prone, and difficult to scale. We present AEAS, an automated system designed to assess and prioritize actionable exploits through static analysis. AEAS analyzes both exploit code and associated documentation to extract a structured set of features reflecting exploit availability, functionality, and setup complexity. It then computes an actionability score for each exploit and produces ranked exploit recommendations. We evaluate AEAS on a dataset of over 5,000 vulnerabilities derived from 600+ real-world applications frequently encountered by red teams. Manual validation and expert review on representative subsets show that AEAS achieves a 100% top-3 success rate in recommending functional exploits and shows strong alignment with expert-validated rankings. These results demonstrate the effectiveness of AEAS in supporting exploit-driven vulnerability prioritization.
Large Language Models (LLMs) struggle with information forgetting and inefficiency in long-horizon, multi-turn dialogues. To address this, we propose a training-free prompt engineering method, the State-Update Multi-turn Dialogue Strategy. It utilizes "State Reconstruction" and "History Remind" mechanisms to effectively manage dialogue history. Our strategy shows strong performance across multiple multi-hop QA datasets. For instance, on the HotpotQA dataset, it improves the core information filtering score by 32.6%, leading to a 14.1% increase in the downstream QA score, while also reducing inference time by 73.1% and token consumption by 59.4%. Ablation studies confirm the pivotal roles of both components. Our work offers an effective solution for optimizing LLMs in long-range interactions, providing new insights for developing more robust Agents.
The evolution of morality presents a puzzle: natural selection should favor self-interest, yet humans developed moral systems promoting altruism. We address this question by introducing a novel Large Language Model (LLM)-based agent simulation framework modeling prehistoric hunter-gatherer societies. This platform is designed to probe diverse questions in social evolution, from survival advantages to inter-group dynamics. To investigate moral evolution, we designed agents with varying moral dispositions based on the Expanding Circle Theory \citep{singer1981expanding}. We evaluated their evolutionary success across a series of simulations and analyzed their decision-making in specially designed moral dilemmas. These experiments reveal how an agent's moral framework, in combination with its cognitive constraints, directly shapes its behavior and determines its evolutionary outcome. Crucially, the emergent patterns echo seminal theories from related domains of social science, providing external validation for the simulations. This work establishes LLM-based simulation as a powerful new paradigm to complement traditional research in evolutionary biology and anthropology, opening new avenues for investigating the complexities of moral and social evolution.
Large Language Models (LLMs) have excelled in question-answering (QA) tasks within single domains. However, their reasoning and coordination capabilities in complex, multi-stage scenarios remain underexplored. Existing benchmarks typically focus on isolated tasks or narrow domains, overlooking models' abilities for multi-stage collaboration and optimization without explicit external guidance. To bridge this gap, we propose \textbf{MSCoRe}, a novel benchmark comprising 126696 domain-specific QA instances spanning scenarios in automotive, pharmaceutical, electronics, and energy sectors. The dataset is created using a structured three-phase pipeline: dynamic sampling, iterative question-answer generation, and a multi-level quality assessment to ensure data quality. Tasks are further categorized into three difficulty levels according to stage coverage and complexity. With MSCoRe, we have conducted a comprehensive evaluation of various state-of-the-art LLM agents. The commercial models performed best across all tasks and scenarios, but a notable gap in ROUGE scores remains between simple and complex tasks. We also tested the models' robustness and found that their performance is negatively affected by noisy data. MSCoRe provides a valuable new resource for the community to evaluate and improve multi-stage reasoning in LLM agents. The code and data are available at https://github.com/D3E0-source/MSCoRE.
Task-oriented conversational systems are essential for efficiently addressing diverse user needs, yet their development requires substantial amounts of high-quality conversational data that is challenging and costly to obtain. While large language models (LLMs) have demonstrated potential in generating synthetic conversations, the extent to which these agent-generated interactions can effectively substitute real human conversations remains unclear. This work presents the first systematic comparison between LLM-simulated users and human users in personalized task-oriented conversations. We propose a comprehensive analytical framework encompassing three key aspects (conversation strategy, interaction style, and conversation evaluation) and ten distinct dimensions for evaluating user behaviors, and collect parallel conversational datasets from both human users and LLM agent users across four representative scenarios under identical conditions. Our analysis reveals significant behavioral differences between the two user types in problem-solving approaches, question broadness, user engagement, context dependency, feedback polarity and promise, language style, and hallucination awareness. We found consistency in the agent users and human users across the depth-first or breadth-first dimensions, as well as the usefulness dimensions. These findings provide critical insights for advancing LLM-based user simulation. Our multi-dimensional taxonomy constructed a generalizable framework for analyzing user behavior patterns, offering insights from LLM agent users and human users. By this work, we provide perspectives on rethinking how to use user simulation in conversational systems in the future.
Large language models (LLMs) have advanced code generation from single-function tasks to competitive-programming problems, but existing multi-agent solutions either rely on costly large-scale ($>$ 30B) models or collapse when downsized to small open-source models. We present MapCoder-Lite, which upgrades a single 7B model into four role-specialised agents-retriever, planner, coder, and debugger-using only rank-32, role-specific LoRA adapters ($<3\%$ extra parameters). Three lightweight techniques make this possible: (i) trajectory distillation from strong LLMs fixes format fragility in retrieval and debugging, (ii) supervisor-guided correction strengthens planning and coding agents, and (iii) agent-wise LoRA fine-tuning delivers memory-efficient specialisation. Comprehensive evaluation on xCodeEval, APPS, and CodeContests shows that MapCoder-Lite more than doubles xCodeEval accuracy (from $13.2\%$ to $28.3\%$), eliminates all format failures, and closes to within six points of a 32B baseline while cutting GPU memory and token-generation time by $4\times$. These results demonstrate that careful agent-wise fine-tuning unleashes high-quality multi-agent coding on a small language model.