Tootfinder

This https://arxiv.org/abs/2505.05849 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csCR_…

AGENTFUZZER: Generic Black-Box Fuzzing for Indirect Prompt Injection against LLM Agents
The strong planning and reasoning capabilities of Large Language Models (LLMs) have fostered the development of agent-based systems capable of leveraging external tools and interacting with increasingly complex environments. However, these powerful features also introduce a critical security risk: indirect prompt injection, a sophisticated attack vector that compromises the core of these agents, the LLM, by manipulating contextual information rather than direct user prompts. In this work, we pr…

Time to Talk: LLM Agents for Asynchronous Group Communication in Mafia Games
Niv Eckhaus, Uri Berger, Gabriel Stanovsky
https://arxiv.org/abs/2506.05309 ht…

Time to Talk: LLM Agents for Asynchronous Group Communication in Mafia Games
LLMs are used predominantly in synchronous communication, where a human user and a model communicate in alternating turns. In contrast, many real-world settings are inherently asynchronous. For example, in group chats, online team meetings, or social games, there is no inherent notion of turns; therefore, the decision of when to speak forms a crucial part of the participant's decision making. In this work, we develop an adaptive asynchronous LLM-agent which, in addition to determining what to s…

Everybody complaining about getting hammered with #AI traffic seems to think that these are crawlers scraping for training data.
How likely is it that this is a complete misconception and this is all inference time?
Most public companies give their cralwers and RAG agents different user agent strings. But what about security services trawling through their data?

This https://arxiv.org/abs/2505.18882 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csCY_…

Personalized Safety in LLMs: A Benchmark and A Planning-Based Agent Approach
Large language models (LLMs) typically generate identical or similar responses for all users given the same prompt, posing serious safety risks in high-stakes applications where user vulnerabilities differ widely. Existing safety evaluations primarily rely on context-independent metrics - such as factuality, bias, or toxicity - overlooking the fact that the same response may carry divergent risks depending on the user's background or condition. We introduce personalized safety to fill this gap …

Quick #CSS tip: If you unset all styles of a button with `all: unset`, don’t forget to re-add all the styles you need, like `box-sizing: border-box` or focus styles, which previously might have been coming from a rule with lower specificity (*) or the user agent stylesheet.

This https://arxiv.org/abs/2502.13172 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csCR_…

Unveiling Privacy Risks in LLM Agent Memory
Large Language Model (LLM) agents have become increasingly prevalent across various real-world applications. They enhance decision-making by storing private user-agent interactions in the memory module for demonstrations, introducing new privacy risks for LLM agents. In this work, we systematically investigate the vulnerability of LLM agents to our proposed Memory EXTRaction Attack (MEXTRA) under a black-box setting. To extract private information from memory, we propose an effective attacking …

Breaking Single-Tester Limits: Multi-Agent LLMs for Multi-User Feature Testing
Sidong Feng, Changhao Du, Huaxiao Liu, Qingnan Wang, Zhengwei Lv, Mengfei Wang, Chunyang Chen
https://arxiv.org/abs/2506.17539

Breaking Single-Tester Limits: Multi-Agent LLMs for Multi-User Feature Testing
The growing dependence on mobile phones and their apps has made multi-user interactive features, like chat calls, live streaming, and video conferencing, indispensable for bridging the gaps in social connectivity caused by physical and situational barriers. However, automating these interactive features for testing is fraught with challenges, owing to their inherent need for timely, dynamic, and collaborative user interactions, which current automated testing methods inadequately address. Inspi…

MAGPIE: A dataset for Multi-AGent contextual PrIvacy Evaluation
Gurusha Juneja, Alon Albalak, Wenyue Hua, William Yang Wang
https://arxiv.org/abs/2506.20737

MAGPIE: A dataset for Multi-AGent contextual PrIvacy Evaluation
The proliferation of LLM-based agents has led to increasing deployment of inter-agent collaboration for tasks like scheduling, negotiation, resource allocation etc. In such systems, privacy is critical, as agents often access proprietary tools and domain-specific databases requiring strict confidentiality. This paper examines whether LLM-based agents demonstrate an understanding of contextual privacy. And, if instructed, do these systems preserve inference time user privacy in non-adversarial m…

SheetMind: An End-to-End LLM-Powered Multi-Agent Framework for Spreadsheet Automation
Ruiyan Zhu, Xi Cheng, Ke Liu, Brian Zhu, Daniel Jin, Neeraj Parihar, Zhoutian Xu, Oliver Gao
https://arxiv.org/abs/2506.12339

SheetMind: An End-to-End LLM-Powered Multi-Agent Framework for Spreadsheet Automation
We present SheetMind, a modular multi-agent framework powered by large language models (LLMs) for spreadsheet automation via natural language instructions. The system comprises three specialized agents: a Manager Agent that decomposes complex user instructions into subtasks; an Action Agent that translates these into structured commands using a Backus Naur Form (BNF) grammar; and a Reflection Agent that validates alignment between generated actions and the user's original intent. Integrated int…

Control at Stake: Evaluating the Security Landscape of LLM-Driven Email Agents
Jiangrong Wu, Yuhong Nan, Jianliang Wu, Zitong Yao, Zibin Zheng
https://arxiv.org/abs/2507.02699

Control at Stake: Evaluating the Security Landscape of LLM-Driven Email Agents
The increasing capabilities of LLMs have led to the rapid proliferation of LLM agent apps, where developers enhance LLMs with access to external resources to support complex task execution. Among these, LLM email agent apps represent one of the widely used categories, as email remains a critical communication medium for users. LLM email agents are capable of managing and responding to email using LLM-driven reasoning and autonomously executing user instructions via external email APIs (e.g., se…

Man könnte ja denken das diese ganze Agent/MCP Geschichte grad doch sehr fast and loose abläuft ....
https://simonwillison.net/2025/May/26/github-mcp-exploited/

ARAG: Agentic Retrieval Augmented Generation for Personalized Recommendation
Reza Yousefi Maragheh, Pratheek Vadla, Priyank Gupta, Kai Zhao, Aysenur Inan, Kehui Yao, Jianpeng Xu, Praveen Kanumala, Jason Cho, Sushant Kumar
https://arxiv.org/abs/2506.21931

ARAG: Agentic Retrieval Augmented Generation for Personalized Recommendation
Retrieval-Augmented Generation (RAG) has shown promise in enhancing recommendation systems by incorporating external context into large language model prompts. However, existing RAG-based approaches often rely on static retrieval heuristics and fail to capture nuanced user preferences in dynamic recommendation scenarios. In this work, we introduce ARAG, an Agentic Retrieval-Augmented Generation framework for Personalized Recommendation, which integrates a multi-agent collaboration mechanism int…

This https://arxiv.org/abs/2501.18160 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csSE_…

RepoAudit: An Autonomous LLM-Agent for Repository-Level Code Auditing
Code auditing is the process of reviewing code with the aim of identifying bugs. Large Language Models (LLMs) have demonstrated promising capabilities for this task without requiring compilation, while also supporting user-friendly customization. However, auditing a code repository with LLMs poses significant challenges: limited context windows and hallucinations can degrade the quality of bug reports, and analyzing large-scale repositories incurs substantial time and token costs, hindering eff…

Researchers find the first known "zero-click" attack on an AI agent; the now-fixed flaw in Microsoft 365 Copilot would let a hacker attack a user via an email (Sharon Goldman/Fortune)
https://fortune.com/2025/06/11/microsoft-cop…

Exclusive: New Microsoft Copilot flaw signals broader risk of AI agents being hacked—‘I would be terrified’
Microsoft fixed the Copilot flaw, but researchers warn the real danger lies in how all AI agents are built.

The content of the user agent is the dumbest thing we've done to the web, and we've doing a whole lot of dumb things to the web—all the way back to spelling referrer incorrectly (and we still do).

LPO: Towards Accurate GUI Agent Interaction via Location Preference Optimization
Jiaqi Tang, Yu Xia, Yi-Feng Wu, Yuwei Hu, Yuhui Chen, Qing-Guo Chen, Xiaogang Xu, Xiangyu Wu, Hao Lu, Yanqing Ma, Shiyin Lu, Qifeng Chen
https://arxiv.org/abs/2506.09373

LPO: Towards Accurate GUI Agent Interaction via Location Preference Optimization
The advent of autonomous agents is transforming interactions with Graphical User Interfaces (GUIs) by employing natural language as a powerful intermediary. Despite the predominance of Supervised Fine-Tuning (SFT) methods in current GUI agents for achieving spatial localization, these methods face substantial challenges due to their limited capacity to accurately perceive positional data. Existing strategies, such as reinforcement learning, often fail to assess positional accuracy effectively, …

I don't recall seeing much measurement research on mobile carrier identification from client traffic. Didn't find much in a cursory search.
Does this exist? Is it feasible, particularly with branded devices like this?
Can classification be done by address pools unique, IDs, MACs, user agent strings, default start-up apps, default DNS queries/settings, etc.?
Clue, ideas, and pointers wanted.

Steve Herman (@w7voa@journa.host)
Attached: 1 image CNBC - A company owned by President Trump announces it own branded Trump Mobile cell phone service and will sell a "T1" smartphone, featuring gold-colored metal case etched with an American flag. https://www.cnbc.com/2025/06/16/trump-mobile-phone-plan.html

🔄 Agent loop architecture: user input − model inference − tool execution − result feedback − continuous conversation flow
💻 String replacement editing approach works effectively with current #AI models for precise code modifications

Joint Beamforming with Extremely Large Scale RIS: A Sequential Multi-Agent A2C Approach
Zhi Chai, Jiajie Xu, Justin P Coon, Mohamed-Slim Alouini
https://arxiv.org/abs/2506.10815

Joint Beamforming with Extremely Large Scale RIS: A Sequential Multi-Agent A2C Approach
It is a challenging problem to jointly optimize the base station (BS) precoding matrix and the reconfigurable intelligent surface (RIS) phases simultaneously in a RIS-assisted multiple-user multiple-input-multiple-output (MU-MIMO) scenario when the size of the RIS becomes extremely large. In this paper, we propose a deep reinforcement learning algorithm called sequential multi-agent advantage actor-critic (A2C) to solve this problem. In addition, the discrete phase of RISs, imperfect channel st…

GUI-Robust: A Comprehensive Dataset for Testing GUI Agent Robustness in Real-World Anomalies
Jingqi Yang, Zhilong Song, Jiawei Chen, Mingli Song, Sheng Zhou, linjun sun, Xiaogang Ouyang, Chun Chen, Can Wang
https://arxiv.org/abs/2506.14477

GUI-Robust: A Comprehensive Dataset for Testing GUI Agent Robustness in Real-World Anomalies
The development of high-quality datasets is crucial for benchmarking and advancing research in Graphical User Interface (GUI) agents. Despite their importance, existing datasets are often constructed under idealized conditions, overlooking the diverse anomalies frequently encountered in real-world deployments. To address this limitation, we introduce GUI-Robust, a novel dataset designed for comprehensive GUI agent evaluation, explicitly incorporating seven common types of anomalies observed in …

MATE: LLM-Powered Multi-Agent Translation Environment for Accessibility Applications
Aleksandr Algazinov, Matt Laing, Paul Laban
https://arxiv.org/abs/2506.19502

MATE: LLM-Powered Multi-Agent Translation Environment for Accessibility Applications
Accessibility remains a critical concern in today's society, as many technologies are not developed to support the full range of user needs. Existing multi-agent systems (MAS) often cannot provide comprehensive assistance for users in need due to the lack of customization stemming from closed-source designs. Consequently, individuals with disabilities frequently encounter significant barriers when attempting to interact with digital environments. We introduce MATE, a multimodal accessibility MA…

Levels of Autonomy for AI Agents
K. J. Kevin Feng, David W. McDonald, Amy X. Zhang
https://arxiv.org/abs/2506.12469 https://arxiv.org…

Levels of Autonomy for AI Agents
Autonomy is a double-edged sword for AI agents, simultaneously unlocking transformative possibilities and serious risks. How can agent developers calibrate the appropriate levels of autonomy at which their agents should operate? We argue that an agent's level of autonomy can be treated as a deliberate design decision, separate from its capability and operational environment. In this work, we define five levels of escalating agent autonomy, characterized by the roles a user can take when interac…

Doppelg\"anger Method: Breaking Role Consistency in LLM Agent via Prompt-based Transferable Adversarial Attack
Daewon Kang, YeongHwan Shin, Doyeon Kim, Kyu-Hwan Jung, Meong Hi Son
https://arxiv.org/abs/2506.14539

Doppelgänger Method: Breaking Role Consistency in LLM Agent via Prompt-based Transferable Adversarial Attack
Since the advent of large language models, prompt engineering now enables the rapid, low-effort creation of diverse autonomous agents that are already in widespread use. Yet this convenience raises urgent concerns about the safety, robustness, and behavioral consistency of the underlying prompts, along with the pressing challenge of preventing those prompts from being exposed to user's attempts. In this paper, we propose the ''Doppelgänger method'' to demonstrate the risk of an agent being hij…

Discovering Coordinated Processes From Social Online Networks
Anna Kalenkova, Lewis Mitchell, Ethan Johnson
https://arxiv.org/abs/2506.12988 https://

Discovering Coordinated Processes From Social Online Networks
The rapid growth of social media presents a unique opportunity to study coordinated agent behavior in an unfiltered environment. Online processes often exhibit complex structures that reflect the nature of the user behavior, whether it is authentic and genuine, or part of a coordinated effort by malicious agents to spread misinformation and disinformation. Detection of AI-generated content can be extremely challenging due to the high quality of large language model-generated text. Therefore, ap…

From Conversation to Orchestration: HCI Challenges and Opportunities in Interactive Multi-Agentic Systems
Sarah Sch\"ombs, Yan Zhang, Jorge Goncalves, Wafa Johal
https://arxiv.org/abs/2506.20091

From Conversation to Orchestration: HCI Challenges and Opportunities in Interactive Multi-Agentic Systems
Recent advances in multi-agentic systems (e.g. AutoGen, OpenAI Swarm) allow users to interact with a group of specialised AI agents rather than a single general-purpose agent. Despite the promise of this new paradigm, the HCI community has yet to fully examine the opportunities, risks, and user-centred challenges it introduces. We contribute to research on multi-agentic systems by exploring their architectures and key features through a human-centred lens. While literature and use cases remain …

Preference-based learning for news headline recommendation
Alexandre Bouras, Audrey Durand, Richard Khoury
https://arxiv.org/abs/2506.06334 https://…

Preference-based learning for news headline recommendation
This study explores strategies for optimizing news headline recommendations through preference-based learning. Using real-world data of user interactions with French-language online news posts, we learn a headline recommender agent under a contextual bandit setting. This allows us to explore the impact of translation on engagement predictions, as well as the benefits of different interactive strategies on user engagement during data collection. Our results show that explicit exploration may not…

Learning, Reasoning, Refinement: A Framework for Kahneman's Dual-System Intelligence in GUI Agents
Jinjie Wei, Jiyao Liu, Lihao Liu, Ming Hu, Junzhi Ning, Mingcheng Li, Weijie Yin, Junjun He, Xiao Liang, Chao Feng, Dingkang Yang
https://arxiv.org/abs/2506.17913

Learning, Reasoning, Refinement: A Framework for Kahneman's Dual-System Intelligence in GUI Agents
Graphical User Interface (GUI) agents have made significant progress in automating digital tasks through the utilization of computer vision and language models. Nevertheless, existing agent systems encounter notable limitations. Firstly, they predominantly depend on trial and error decision making rather than progressive reasoning, thereby lacking the capability to learn and adapt from interactive encounters. Secondly, these systems are assessed using overly simplistic single step accuracy metr…

StorySage: Conversational Autobiography Writing Powered by a Multi-Agent Framework
Shayan Talaei, Meijin Li, Kanu Grover, James Kent Hippler, Diyi Yang, Amin Saberi
https://arxiv.org/abs/2506.14159

StorySage: Conversational Autobiography Writing Powered by a Multi-Agent Framework
Every individual carries a unique and personal life story shaped by their memories and experiences. However, these memories are often scattered and difficult to organize into a coherent narrative, a challenge that defines the task of autobiography writing. Existing conversational writing assistants tend to rely on generic user interactions and pre-defined guidelines, making it difficult for these systems to capture personal memories and develop a complete biography over time. We introduce Story…

IntenTest: Stress Testing for Intent Integrity in API-Calling LLM Agents
Shiwei Feng, Xiangzhe Xu, Xuan Chen, Kaiyuan Zhang, Syed Yusuf Ahmed, Zian Su, Mingwei Zheng, Xiangyu Zhang
https://arxiv.org/abs/2506.07524

IntenTest: Stress Testing for Intent Integrity in API-Calling LLM Agents
LLM agents are increasingly deployed to automate real-world tasks by invoking APIs through natural language instructions. While powerful, they often suffer from misinterpretation of user intent, leading to the agent's actions that diverge from the user's intended goal, especially as external toolkits evolve. Traditional software testing assumes structured inputs and thus falls short in handling the ambiguity of natural language. We introduce IntenTest, an API-centric stress testing framework th…

Effective Red-Teaming of Policy-Adherent Agents
Itay Nakash, George Kour, Koren Lazar, Matan Vetzler, Guy Uziel, Ateret Anaby-Tavor
https://arxiv.org/abs/2506.09600

Effective Red-Teaming of Policy-Adherent Agents
Task-oriented LLM-based agents are increasingly used in domains with strict policies, such as refund eligibility or cancellation rules. The challenge lies in ensuring that the agent consistently adheres to these rules and policies, appropriately refusing any request that would violate them, while still maintaining a helpful and natural interaction. This calls for the development of tailored design and evaluation methodologies to ensure agent resilience against malicious user behavior. We propos…

Affective-CARA: A Knowledge Graph Driven Framework for Culturally Adaptive Emotional Intelligence in HCI
Nirodya Pussadeniya, Bahareh Nakisa, Mohmmad Naim Rastgoo
https://arxiv.org/abs/2506.14166

Affective-CARA: A Knowledge Graph Driven Framework for Culturally Adaptive Emotional Intelligence in HCI
Culturally adaptive emotional responses remain a critical challenge in affective computing. This paper introduces Affective-CARA, an agentic framework designed to enhance user-agent interactions by integrating a Cultural Emotion Knowledge Graph (derived from StereoKG) with Valence, Arousal, and Dominance annotations, culture-specific data, and cross-cultural checks to minimize bias. A Gradient-Based Reward Policy Optimization mechanism further refines responses according to cultural alignment, …

Understanding Software Engineering Agents Through the Lens of Traceability: An Empirical Study
Ira Ceka, Saurabh Pujar, Shyam Ramji, Luca Buratti, Gail Kaiser, Baishakhi Ray
https://arxiv.org/abs/2506.08311

Understanding Software Engineering Agents Through the Lens of Traceability: An Empirical Study
With the advent of large language models (LLMs), software engineering agents (SWE agents) have emerged as a powerful paradigm for automating a range of software tasks -- from code generation and repair to test case synthesis. These agents operate autonomously by interpreting user input and responding to environmental feedback. While various agent architectures have demonstrated strong empirical performance, the internal decision-making worfklows that drive their behavior remain poorly understoo…

This https://arxiv.org/abs/2505.05660 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csHC_…

Not Like Us, Hunty: Measuring Perceptions and Behavioral Effects of Minoritized Anthropomorphic Cues in LLMs
As large language models (LLMs) increasingly adapt and personalize to diverse sets of users, there is an increased risk of systems appropriating sociolects, i.e., language styles or dialects that are associated with specific minoritized lived experiences (e.g., African American English, Queer slang). In this work, we examine whether sociolect usage by an LLM agent affects user reliance on its outputs and user perception (satisfaction, frustration, trust, and social presence). We designed and co…