Despite the transformative impact of Artificial Intelligence (AI) across various sectors, cyber security continues to rely on traditional static and dynamic analysis tools, hampered by high false positive rates and superficial code comprehension. While generative AI offers promising automation capabilities for software development, leveraging Large Language Models (LLMs) for vulnerability detection presents unique challenges. This paper explores the potential and limitations of LLMs in identifying vulnerabilities, acknowledging inherent weaknesses such as hallucinations, limited context length, and knowledge cut-offs. Previous attempts employing machine learning models for vulnerability detection have proven ineffective due to limited real-world applicability, feature engineering challenges, lack of contextual understanding, and the complexities of training models to keep pace with the evolving threat landscape. Therefore, we propose a robust AI-driven approach focused on mitigating these limitations and ensuring the quality and reliability of LLM based vulnerability detection. Through innovative methodologies combining Retrieval-Augmented Generation (RAG) and Mixtureof-Agents (MoA), this research seeks to leverage the strengths of LLMs while addressing their weaknesses, ultimately paving the way for dependable and efficient AI-powered solutions in securing the ever-evolving software landscape.
In recent years, multi-agent frameworks powered by large language models (LLMs) have advanced rapidly. Despite this progress, there is still a notable absence of benchmark datasets specifically tailored to evaluate their performance. To bridge this gap, we introduce Auto-SLURP, a benchmark dataset aimed at evaluating LLM-based multi-agent frameworks in the context of intelligent personal assistants. Auto-SLURP extends the original SLURP dataset -- initially developed for natural language understanding tasks -- by relabeling the data and integrating simulated servers and external services. This enhancement enables a comprehensive end-to-end evaluation pipeline, covering language understanding, task execution, and response generation. Our experiments demonstrate that Auto-SLURP presents a significant challenge for current state-of-the-art frameworks, highlighting that truly reliable and intelligent multi-agent personal assistants remain a work in progress. The dataset and related code are available at https://github.com/lorashen/Auto-SLURP/.
With the surge of large language models (LLMs), Large Vision-Language Models (VLMs)--which integrate vision encoders with LLMs for accurate visual grounding--have shown great potential in tasks like generalist agents and robotic control. However, VLMs are typically trained on massive web-scraped images, raising concerns over copyright infringement and privacy violations, and making data auditing increasingly urgent. Membership inference (MI), which determines whether a sample was used in training, has emerged as a key auditing technique, with promising results on open-source VLMs like LLaVA (AUC > 80%). In this work, we revisit these advances and uncover a critical issue: current MI benchmarks suffer from distribution shifts between member and non-member images, introducing shortcut cues that inflate MI performance. We further analyze the nature of these shifts and propose a principled metric based on optimal transport to quantify the distribution discrepancy. To evaluate MI in realistic settings, we construct new benchmarks with i.i.d. member and non-member images. Existing MI methods fail under these unbiased conditions, performing only marginally better than chance. Further, we explore the theoretical upper bound of MI by probing the Bayes Optimality within the VLM's embedding space and find the irreducible error rate remains high. Despite this pessimistic outlook, we analyze why MI for VLMs is particularly challenging and identify three practical scenarios--fine-tuning, access to ground-truth texts, and set-based inference--where auditing becomes feasible. Our study presents a systematic view of the limits and opportunities of MI for VLMs, providing guidance for future efforts in trustworthy data auditing.
Using multiple agents was found to improve the debugging capabilities of Large Language Models. However, increasing the number of LLM-agents has several drawbacks such as increasing the running costs and rising the risk for the agents to lose focus. In this work, we propose an adaptive agentic design, where the number of agents and their roles are determined dynamically based on the characteristics of the task to be achieved. In this design, the agents roles are not predefined, but are generated after analyzing the problem to be solved. Our initial evaluation shows that, with the adaptive design, the number of agents that are generated depends on the complexity of the buggy code. In fact, for simple code with mere syntax issues, the problem was usually fixed using one agent only. However, for more complex problems, we noticed the creation of a higher number of agents. Regarding the effectiveness of the fix, we noticed an average improvement of 11% compared to the one-shot prompting. Given these promising results, we outline future research directions to improve our design for adaptive software agents that can autonomously plan and conduct their software goals.
Automating structured clinical interviews could revolutionize mental healthcare accessibility, yet existing large language models (LLMs) approaches fail to align with psychiatric diagnostic protocols. We present MAGI, the first framework that transforms the gold-standard Mini International Neuropsychiatric Interview (MINI) into automatic computational workflows through coordinated multi-agent collaboration. MAGI dynamically navigates clinical logic via four specialized agents: 1) an interview tree guided navigation agent adhering to the MINI's branching structure, 2) an adaptive question agent blending diagnostic probing, explaining, and empathy, 3) a judgment agent validating whether the response from participants meet the node, and 4) a diagnosis Agent generating Psychometric Chain-of- Thought (PsyCoT) traces that explicitly map symptoms to clinical criteria. Experimental results on 1,002 real-world participants covering depression, generalized anxiety, social anxiety and suicide shows that MAGI advances LLM- assisted mental health assessment by combining clinical rigor, conversational adaptability, and explainable reasoning.
As modern vehicles evolve into intelligent and connected systems, their growing complexity introduces significant cybersecurity risks. Threat Analysis and Risk Assessment (TARA) has therefore become essential for managing these risks under mandatory regulations. However, existing TARA automation methods rely on static threat libraries, limiting their utility in the detailed, function-level analyses demanded by industry. This paper introduces DefenseWeaver, the first system that automates function-level TARA using component-specific details and large language models (LLMs). DefenseWeaver dynamically generates attack trees and risk evaluations from system configurations described in an extended OpenXSAM++ format, then employs a multi-agent framework to coordinate specialized LLM roles for more robust analysis. To further adapt to evolving threats and diverse standards, DefenseWeaver incorporates Low-Rank Adaptation (LoRA) fine-tuning and Retrieval-Augmented Generation (RAG) with expert-curated TARA reports. We validated DefenseWeaver through deployment in four automotive security projects, where it identified 11 critical attack paths, verified through penetration testing, and subsequently reported and remediated by the relevant automakers and suppliers. Additionally, DefenseWeaver demonstrated cross-domain adaptability, successfully applying to unmanned aerial vehicles (UAVs) and marine navigation systems. In comparison to human experts, DefenseWeaver outperformed manual attack tree generation across six assessment scenarios. Integrated into commercial cybersecurity platforms such as UAES and Xiaomi, DefenseWeaver has generated over 8,200 attack trees. These results highlight its ability to significantly reduce processing time, and its scalability and transformative impact on cybersecurity across industries.
Large Language Model (LLM) agents have demonstrated impressive capabilities in social deduction games (SDGs) like Werewolf, where strategic reasoning and social deception are essential. However, current approaches remain limited to textual information, ignoring crucial multimodal cues such as facial expressions and tone of voice that humans naturally use to communicate. Moreover, existing SDG agents primarily focus on inferring other players' identities without modeling how others perceive themselves or fellow players. To address these limitations, we use One Night Ultimate Werewolf (ONUW) as a testbed and present MultiMind, the first framework integrating multimodal information into SDG agents. MultiMind processes facial expressions and vocal tones alongside verbal content, while employing a Theory of Mind (ToM) model to represent each player's suspicion levels toward others. By combining this ToM model with Monte Carlo Tree Search (MCTS), our agent identifies communication strategies that minimize suspicion directed at itself. Through comprehensive evaluation in both agent-versus-agent simulations and studies with human players, we demonstrate MultiMind's superior performance in gameplay. Our work presents a significant advancement toward LLM agents capable of human-like social reasoning across multimodal domains.
Drug discovery remains a formidable challenge: more than 90 percent of candidate molecules fail in clinical evaluation, and development costs often exceed one billion dollars per approved therapy. Disparate data streams, from genomics and transcriptomics to chemical libraries and clinical records, hinder coherent mechanistic insight and slow progress. Meanwhile, large language models excel at reasoning and tool integration but lack the modular specialization and iterative memory required for regulated, hypothesis-driven workflows. We introduce PharmaSwarm, a unified multi-agent framework that orchestrates specialized LLM "agents" to propose, validate, and refine hypotheses for novel drug targets and lead compounds. Each agent accesses dedicated functionality--automated genomic and expression analysis; a curated biomedical knowledge graph; pathway enrichment and network simulation; interpretable binding affinity prediction--while a central Evaluator LLM continuously ranks proposals by biological plausibility, novelty, in silico efficacy, and safety. A shared memory layer captures validated insights and fine-tunes underlying submodels over time, yielding a self-improving system. Deployable on low-code platforms or Kubernetes-based microservices, PharmaSwarm supports literature-driven discovery, omics-guided target identification, and market-informed repurposing. We also describe a rigorous four-tier validation pipeline spanning retrospective benchmarking, independent computational assays, experimental testing, and expert user studies to ensure transparency, reproducibility, and real-world impact. By acting as an AI copilot, PharmaSwarm can accelerate translational research and deliver high-confidence hypotheses more efficiently than traditional pipelines.
Collaboration is ubiquitous and essential in day-to-day life -- from exchanging ideas, to delegating tasks, to generating plans together. This work studies how LLMs can adaptively collaborate to perform complex embodied reasoning tasks. To this end we introduce MINDcraft, an easily extensible platform built to enable LLM agents to control characters in the open-world game of Minecraft; and MineCollab, a benchmark to test the different dimensions of embodied and collaborative reasoning. An experimental study finds that the primary bottleneck in collaborating effectively for current state-of-the-art agents is efficient natural language communication, with agent performance dropping as much as 15% when they are required to communicate detailed task completion plans. We conclude that existing LLM agents are ill-optimized for multi-agent collaboration, especially in embodied scenarios, and highlight the need to employ methods beyond in-context and imitation learning. Our website can be found here: https://mindcraft-minecollab.github.io/
The rise of Large Language Models (LLMs) has revolutionized Graphical User Interface (GUI) automation through LLM-powered GUI agents, yet their ability to process sensitive data with limited human oversight raises significant privacy and security risks. This position paper identifies three key risks of GUI agents and examines how they differ from traditional GUI automation and general autonomous agents. Despite these risks, existing evaluations focus primarily on performance, leaving privacy and security assessments largely unexplored. We review current evaluation metrics for both GUI and general LLM agents and outline five key challenges in integrating human evaluators for GUI agent assessments. To address these gaps, we advocate for a human-centered evaluation framework that incorporates risk assessments, enhances user awareness through in-context consent, and embeds privacy and security considerations into GUI agent design and evaluation.
Agentic AI systems powered by Large Language Models (LLMs) as their foundational reasoning engine, are transforming clinical workflows such as medical report generation and clinical summarization by autonomously analyzing sensitive healthcare data and executing decisions with minimal human oversight. However, their adoption demands strict compliance with regulatory frameworks such as Health Insurance Portability and Accountability Act (HIPAA), particularly when handling Protected Health Information (PHI). This work-in-progress paper introduces a HIPAA-compliant Agentic AI framework that enforces regulatory compliance through dynamic, context-aware policy enforcement. Our framework integrates three core mechanisms: (1) Attribute-Based Access Control (ABAC) for granular PHI governance, (2) a hybrid PHI sanitization pipeline combining regex patterns and BERT-based model to minimize leakage, and (3) immutable audit trails for compliance verification.
Feature selection aims to preprocess the target dataset, find an optimal and most streamlined feature subset, and enhance the downstream machine learning task. Among filter, wrapper, and embedded-based approaches, the reinforcement learning (RL)-based subspace exploration strategy provides a novel objective optimization-directed perspective and promising performance. Nevertheless, even with improved performance, current reinforcement learning approaches face challenges similar to conventional methods when dealing with complex datasets. These challenges stem from the inefficient paradigm of using one agent per feature and the inherent complexities present in the datasets. This observation motivates us to investigate and address the above issue and propose a novel approach, namely HRLFS. Our methodology initially employs a Large Language Model (LLM)-based hybrid state extractor to capture each feature's mathematical and semantic characteristics. Based on this information, features are clustered, facilitating the construction of hierarchical agents for each cluster and sub-cluster. Extensive experiments demonstrate the efficiency, scalability, and robustness of our approach. Compared to contemporary or the one-feature-one-agent RL-based approaches, HRLFS improves the downstream ML performance with iterative feature subspace exploration while accelerating total run time by reducing the number of agents involved.
A data story typically integrates data facts from multiple perspectives and stances to construct a comprehensive and objective narrative. However, retrieving these facts demands time for data search and challenges the creator's analytical skills. In this work, we introduce DataScout, an interactive system that automatically performs reasoning and stance-based data facts retrieval to augment the user's statement. Particularly, DataScout leverages an LLM-based agent to construct a retrieval tree, enabling collaborative control of its expansion between users and the agent. The interface visualizes the retrieval tree as a mind map that eases users to intuitively steer the retrieval direction and effectively engage in reasoning and analysis. We evaluate the proposed system through case studies and in-depth expert interviews. Our evaluation demonstrates that DataScout can effectively retrieve multifaceted data facts from different stances, helping users verify their statements and enhance the credibility of their stories.
Large language models (LLMs) are a transformational capability at the frontier of artificial intelligence and machine learning that can support decision-makers in addressing pressing societal challenges such as extreme natural hazard events. As generalized models, LLMs often struggle to provide context-specific information, particularly in areas requiring specialized knowledge. In this work we propose a retrieval-augmented generation (RAG)-based multi-agent LLM system to support analysis and decision-making in the context of natural hazards and extreme weather events. As a proof of concept, we present WildfireGPT, a specialized system focused on wildfire hazards. The architecture employs a user-centered, multi-agent design to deliver tailored risk insights across diverse stakeholder groups. By integrating natural hazard and extreme weather projection data, observational datasets, and scientific literature through an RAG framework, the system ensures both the accuracy and contextual relevance of the information it provides. Evaluation across ten expert-led case studies demonstrates that WildfireGPT significantly outperforms existing LLM-based solutions for decision support.
Despite the rapid growth of machine learning research, corresponding code implementations are often unavailable, making it slow and labor-intensive for researchers to reproduce results and build upon prior work. In the meantime, recent Large Language Models (LLMs) excel at understanding scientific documents and generating high-quality code. Inspired by this, we introduce PaperCoder, a multi-agent LLM framework that transforms machine learning papers into functional code repositories. PaperCoder operates in three stages: planning, where it constructs a high-level roadmap, designs the system architecture with diagrams, identifies file dependencies, and generates configuration files; analysis, which focuses on interpreting implementation-specific details; and generation, where modular, dependency-aware code is produced. Moreover, each phase is instantiated through a set of specialized agents designed to collaborate effectively across the pipeline. We then evaluate PaperCoder on generating code implementations from machine learning papers based on both model-based and human evaluations, specifically from the original paper authors, with author-released repositories as ground truth if available. Our results demonstrate the effectiveness of PaperCoder in creating high-quality, faithful implementations. Furthermore, it consistently shows strengths in the recently released PaperBench benchmark, surpassing strong baselines by substantial margins.
Large language models (LLMs) are being widely applied across various fields, but as tasks become more complex, evaluating their responses is increasingly challenging. Compared to human evaluators, the use of LLMs to support performance evaluation offers a more efficient alternative. However, most studies focus mainly on aligning LLMs' judgments with human preferences, overlooking the existence of biases and mistakes in human judgment. Furthermore, how to select suitable LLM judgments given multiple potential LLM responses remains underexplored. To address these two aforementioned issues, we propose a three-stage meta-judge selection pipeline: 1) developing a comprehensive rubric with GPT-4 and human experts, 2) using three advanced LLM agents to score judgments, and 3) applying a threshold to filter out low-scoring judgments. Compared to methods using a single LLM as both judge and meta-judge, our pipeline introduces multi-agent collaboration and a more comprehensive rubric. Experimental results on the JudgeBench dataset show about 15.55\% improvement compared to raw judgments and about 8.37\% improvement over the single-agent baseline. Our work demonstrates the potential of LLMs as meta-judges and lays the foundation for future research on constructing preference datasets for LLM-as-a-judge reinforcement learning.
Optimization plays a vital role in scientific research and practical applications, but formulating a concrete optimization problem described in natural language into a mathematical form and selecting a suitable solver to solve the problem requires substantial domain expertise. We introduce \textbf{OptimAI}, a framework for solving \underline{Optim}ization problems described in natural language by leveraging LLM-powered \underline{AI} agents, achieving superior performance over current state-of-the-art methods. Our framework is built upon four key roles: (1) a \emph{formulator} that translates natural language problem descriptions into precise mathematical formulations; (2) a \emph{planner} that constructs a high-level solution strategy prior to execution; and (3) a \emph{coder} and a \emph{code critic} capable of interacting with the environment and reflecting on outcomes to refine future actions. Ablation studies confirm that all roles are essential; removing the planner or code critic results in $5.8\times$ and $3.1\times$ drops in productivity, respectively. Furthermore, we introduce UCB-based debug scheduling to dynamically switch between alternative plans, yielding an additional $3.3\times$ productivity gain. Our design emphasizes multi-agent collaboration, allowing us to conveniently explore the synergistic effect of combining diverse models within a unified system. Our approach attains 88.1\% accuracy on the NLP4LP dataset and 71.2\% on the Optibench (non-linear w/o table) subset, reducing error rates by 58\% and 50\% respectively over prior best results.
Large Language Models (LLMs) are commonly criticized for not understanding language. However, many critiques focus on cognitive abilities that, in humans, are distinct from language processing. Here, we instead study a kind of understanding tightly linked to language: inferring who did what to whom (thematic roles) in a sentence. Does the central training objective of LLMs-word prediction-result in sentence representations that capture thematic roles? In two experiments, we characterized sentence representations in four LLMs. In contrast to human similarity judgments, in LLMs the overall representational similarity of sentence pairs reflected syntactic similarity but not whether their agent and patient assignments were identical vs. reversed. Furthermore, we found little evidence that thematic role information was available in any subset of hidden units. However, some attention heads robustly captured thematic roles, independently of syntax. Therefore, LLMs can extract thematic roles but, relative to humans, this information influences their representations more weakly.
Text-based games provide valuable environments for language-based autonomous agents. However, planning-then-learning paradigms, such as those combining Monte Carlo Tree Search (MCTS) and reinforcement learning (RL), are notably time-consuming due to extensive iterations. Additionally, these algorithms perform uncertainty-driven exploration but lack language understanding and reasoning abilities. In this paper, we introduce the Monte Carlo planning with Dynamic Memory-guided Large language model (MC-DML) algorithm. MC-DML leverages the language understanding and reasoning capabilities of Large Language Models (LLMs) alongside the exploratory advantages of tree search algorithms. Specifically, we enhance LLMs with in-trial and cross-trial memory mechanisms, enabling them to learn from past experiences and dynamically adjust action evaluations during planning. We conduct experiments on a series of text-based games from the Jericho benchmark. Our results demonstrate that the MC-DML algorithm significantly enhances performance across various games at the initial planning phase, outperforming strong contemporary methods that require multiple iterations. This demonstrates the effectiveness of our algorithm, paving the way for more efficient language-grounded planning in complex environments.
The rapid development of large language models (LLMs) has led to the widespread deployment of LLM agents across diverse industries, including customer service, content generation, data analysis, and even healthcare. However, as more LLM agents are deployed, a major issue has emerged: there is no standard way for these agents to communicate with external tools or data sources. This lack of standardized protocols makes it difficult for agents to work together or scale effectively, and it limits their ability to tackle complex, real-world tasks. A unified communication protocol for LLM agents could change this. It would allow agents and tools to interact more smoothly, encourage collaboration, and triggering the formation of collective intelligence. In this paper, we provide a systematic overview of existing communication protocols for LLM agents. We classify them into four main categories and make an analysis to help users and developers select the most suitable protocols for specific applications. Additionally, we conduct a comparative performance analysis of these protocols across key dimensions such as security, scalability, and latency. Finally, we explore future challenges, such as how protocols can adapt and survive in fast-evolving environments, and what qualities future protocols might need to support the next generation of LLM agent ecosystems. We expect this work to serve as a practical reference for both researchers and engineers seeking to design, evaluate, or integrate robust communication infrastructures for intelligent agents.
The rapid advancement in capabilities of large language models (LLMs) raises a pivotal question: How can LLMs accelerate scientific discovery? This work tackles the crucial first stage of research, generating novel hypotheses. While recent work on automated hypothesis generation focuses on multi-agent frameworks and extending test-time compute, none of the approaches effectively incorporate transparency and steerability through a synergistic Human-in-the-loop (HITL) approach. To address this gap, we introduce IRIS: Interactive Research Ideation System, an open-source platform designed for researchers to leverage LLM-assisted scientific ideation. IRIS incorporates innovative features to enhance ideation, including adaptive test-time compute expansion via Monte Carlo Tree Search (MCTS), fine-grained feedback mechanism, and query-based literature synthesis. Designed to empower researchers with greater control and insight throughout the ideation process. We additionally conduct a user study with researchers across diverse disciplines, validating the effectiveness of our system in enhancing ideation. We open-source our code at https://github.com/Anikethh/IRIS-Interactive-Research-Ideation-System
Intelligent agent systems based on Large Language Models (LLMs) have shown great potential in real-world applications. However, existing agent frameworks still face critical limitations in task planning and execution, restricting their effectiveness and generalizability. Specifically, current planning methods often lack clear global goals, leading agents to get stuck in local branches, or produce non-executable plans. Meanwhile, existing execution mechanisms struggle to balance complexity and stability, and their limited action space restricts their ability to handle diverse real-world tasks. To address these limitations, we propose GoalAct, a novel agent framework that introduces a continuously updated global planning mechanism and integrates a hierarchical execution strategy. GoalAct decomposes task execution into high-level skills, including searching, coding, writing and more, thereby reducing planning complexity while enhancing the agents' adaptability across diverse task scenarios. We evaluate GoalAct on LegalAgentBench, a benchmark with multiple types of legal tasks that require the use of multiple types of tools. Experimental results demonstrate that GoalAct achieves state-of-the-art (SOTA) performance, with an average improvement of 12.22% in success rate. These findings highlight GoalAct's potential to drive the development of more advanced intelligent agent systems, making them more effective across complex real-world applications. Our code can be found at https://github.com/cjj826/GoalAct.
Multi-Agent Debate (MAD), leveraging collaborative interactions among Large Language Models (LLMs), aim to enhance reasoning capabilities in complex tasks. However, the security implications of their iterative dialogues and role-playing characteristics, particularly susceptibility to jailbreak attacks eliciting harmful content, remain critically underexplored. This paper systematically investigates the jailbreak vulnerabilities of four prominent MAD frameworks built upon leading commercial LLMs (GPT-4o, GPT-4, GPT-3.5-turbo, and DeepSeek) without compromising internal agents. We introduce a novel structured prompt-rewriting framework specifically designed to exploit MAD dynamics via narrative encapsulation, role-driven escalation, iterative refinement, and rhetorical obfuscation. Our extensive experiments demonstrate that MAD systems are inherently more vulnerable than single-agent setups. Crucially, our proposed attack methodology significantly amplifies this fragility, increasing average harmfulness from 28.14% to 80.34% and achieving attack success rates as high as 80% in certain scenarios. These findings reveal intrinsic vulnerabilities in MAD architectures and underscore the urgent need for robust, specialized defenses prior to real-world deployment.
The XLLM@ACL2025 Shared Task-III formulates a low-resource structural reasoning task that challenges LLMs to generate interpretable, step-by-step rationales with minimal labeled data. We present Less is More, the third-place winning approach in the XLLM@ACL2025 Shared Task-III, which focuses on structured reasoning from only 24 labeled examples. Our approach leverages a multi-agent framework with reverse-prompt induction, retrieval-augmented reasoning synthesis via GPT-4o, and dual-stage reward-guided filtering to distill high-quality supervision across three subtasks: question parsing, CoT parsing, and step-level verification. All modules are fine-tuned from Meta-Llama-3-8B-Instruct under a unified LoRA+ setup. By combining structure validation with reward filtering across few-shot and zero-shot prompts, our pipeline consistently improves structure reasoning quality. These results underscore the value of controllable data distillation in enhancing structured inference under low-resource constraints. Our code is available at https://github.com/Jiahao-Yuan/Less-is-More.
Large language models (LLMs) have demonstrated remarkable capabilities in code generation tasks. However, a significant gap remains between their current performance and that of expert software engineers. A key differentiator is that human engineers actively seek clarification when faced with ambiguous requirements, while LLMs typically generate code regardless of uncertainties in the problem description. We present ClarifyCoder, a novel framework with synthetic data generation and instruction-tuning that enables LLMs to identify ambiguities and request clarification before proceeding with code generation. While recent work has focused on LLM-based agents for iterative code generation, we argue that the fundamental ability to recognize and query ambiguous requirements should be intrinsic to the models themselves. Our approach consists of two main components: (1) a data synthesis technique that augments existing programming datasets with scenarios requiring clarification to generate clarification-aware training data, and (2) a fine-tuning strategy that teaches models to prioritize seeking clarification over immediate code generation when faced with incomplete or ambiguous requirements. We further provide an empirical analysis of integrating ClarifyCoder with standard fine-tuning for a joint optimization of both clarify-awareness and coding ability. Experimental results demonstrate that ClarifyCoder significantly improves the communication capabilities of Code LLMs through meaningful clarification dialogues while maintaining code generation capabilities.
We propose MR. Video, an agentic long video understanding framework that demonstrates the simple yet effective MapReduce principle for processing long videos: (1) Map: independently and densely perceiving short video clips, and (2) Reduce: jointly aggregating information from all clips. Compared with sequence-to-sequence vision-language models (VLMs), MR. Video performs detailed short video perception without being limited by context length. Compared with existing video agents that typically rely on sequential key segment selection, the Map operation enables simpler and more scalable sequence parallel perception of short video segments. Its Reduce step allows for more comprehensive context aggregation and reasoning, surpassing explicit key segment retrieval. This MapReduce principle is applicable to both VLMs and video agents, and we use LLM agents to validate its effectiveness. In practice, MR. Video employs two MapReduce stages: (A) Captioning: generating captions for short video clips (map), then standardizing repeated characters and objects into shared names (reduce); (B) Analysis: for each user question, analyzing relevant information from individual short videos (map), and integrating them into a final answer (reduce). MR. Video achieves over 10% accuracy improvement on the challenging LVBench compared to state-of-the-art VLMs and video agents. Code is available at: https://github.com/ziqipang/MR-Video
The success of Large Language Models (LLMs) has sparked interest in various agentic applications. A key hypothesis is that LLMs, leveraging common sense and Chain-of-Thought (CoT) reasoning, can effectively explore and efficiently solve complex domains. However, LLM agents have been found to suffer from sub-optimal exploration and the knowing-doing gap, the inability to effectively act on knowledge present in the model. In this work, we systematically study why LLMs perform sub-optimally in decision-making scenarios. In particular, we closely examine three prevalent failure modes: greediness, frequency bias, and the knowing-doing gap. We propose mitigation of these shortcomings by fine-tuning via Reinforcement Learning (RL) on self-generated CoT rationales. Our experiments across multi-armed bandits, contextual bandits, and Tic-tac-toe, demonstrate that RL fine-tuning enhances the decision-making abilities of LLMs by increasing exploration and narrowing the knowing-doing gap. Finally, we study both classic exploration mechanisms, such as $\epsilon$-greedy, and LLM-specific approaches, such as self-correction and self-consistency, to enable more effective fine-tuning of LLMs for decision-making.
In Android GUI testing, generating an action sequence for a task that can be replayed as a test script is common. Generating sequences of actions and respective test scripts from task goals described in natural language can eliminate the need for manually writing test scripts. However, existing approaches based on large language models (LLM) often struggle with identifying the final action, and either end prematurely or continue past the final screen. In this paper, we introduce VisiDroid, a multi-modal, LLM-based, multi-agent framework that iteratively determines the next action and leverages visual images of screens to detect the task's completeness. The multi-modal approach enhances our model in two significant ways. First, this approach enables it to avoid prematurely terminating a task when textual content alone provides misleading indications of task completion. Additionally, visual input helps the tool avoid errors when changes in the GUI do not directly affect functionality toward task completion, such as adjustments to font sizes or colors. Second, the multi-modal approach also ensures the tool not progress beyond the final screen, which might lack explicit textual indicators of task completion but could display a visual element indicating task completion, which is common in GUI apps. Our evaluation shows that VisiDroid achieves an accuracy of 87.3%, outperforming the best baseline relatively by 23.5%. We also demonstrate that our multi-modal framework with images and texts enables the LLM to better determine when a task is completed.
As large language models (LLMs) and LLM-based agents increasingly interact with humans in decision-making contexts, understanding the trust dynamics between humans and AI agents becomes a central concern. While considerable literature studies how humans trust AI agents, it is much less understood how LLM-based agents develop effective trust in humans. LLM-based agents likely rely on some sort of implicit effective trust in trust-related contexts (e.g., evaluating individual loan applications) to assist and affect decision making. Using established behavioral theories, we develop an approach that studies whether LLMs trust depends on the three major trustworthiness dimensions: competence, benevolence and integrity of the human subject. We also study how demographic variables affect effective trust. Across 43,200 simulated experiments, for five popular language models, across five different scenarios we find that LLM trust development shows an overall similarity to human trust development. We find that in most, but not all cases, LLM trust is strongly predicted by trustworthiness, and in some cases also biased by age, religion and gender, especially in financial scenarios. This is particularly true for scenarios common in the literature and for newer models. While the overall patterns align with human-like mechanisms of effective trust formation, different models exhibit variation in how they estimate trust; in some cases, trustworthiness and demographic factors are weak predictors of effective trust. These findings call for a better understanding of AI-to-human trust dynamics and monitoring of biases and trust development patterns to prevent unintended and potentially harmful outcomes in trust-sensitive applications of AI.
Can we build accurate world models out of large language models (LLMs)? How can world models benefit LLM agents? The gap between the prior knowledge of LLMs and the specified environment's dynamics usually bottlenecks LLMs' performance as world models. To bridge the gap, we propose a training-free "world alignment" that learns an environment's symbolic knowledge complementary to LLMs. The symbolic knowledge covers action rules, knowledge graphs, and scene graphs, which are extracted by LLMs from exploration trajectories and encoded into executable codes to regulate LLM agents' policies. We further propose an RL-free, model-based agent "WALL-E 2.0" through the model-predictive control (MPC) framework. Unlike classical MPC requiring costly optimization on the fly, we adopt an LLM agent as an efficient look-ahead optimizer of future steps' actions by interacting with the neurosymbolic world model. While the LLM agent's strong heuristics make it an efficient planner in MPC, the quality of its planned actions is also secured by the accurate predictions of the aligned world model. They together considerably improve learning efficiency in a new environment. On open-world challenges in Mars (Minecraft like) and ALFWorld (embodied indoor environments), WALL-E 2.0 significantly outperforms existing methods, e.g., surpassing baselines in Mars by 16.1%-51.6% of success rate and by at least 61.7% in score. In ALFWorld, it achieves a new record 98% success rate after only 4 iterations.