1 citations · 1 across the 13 of their papers we have counts for
20 papers
Benchmarks Are Surprisingly Compaction-Friendly
Jeanne Salle, Hardik Bhatnagar, Matthias Bethge +2
Long horizon agentic tasks are now pushing LLM agents past their context window. To continue the task, the agent scaffold has to free up the context window, a process known as cont…
SCARCE: Emergent Institutions in Long-Horizon Multi-Agent Coding Systems
Abhinav Kumar, Rediet Abebe, Eugene Bagdasarian +1
As autonomous agents become widely deployed, coordinating access to shared computing resources and services will become increasingly important. Recent incidents have already reveal…
Surviving the Router: Optimizing Skill Injections for Retrieval and Execution
Haneen Najjar, Luca Scionis, Haritz Puerto +1
AI agents increasingly rely on modular third-party "skills" that are dynamically selected by skill routers to execute complex tasks. While recent studies highlight the threat of pr…
FIGS: Evaluating Multi-Turn Sycophancy Without Penalizing Empathy
Sidharth Pulipaka, Ruta Binkyte, Ivaxi Sheth +1
Large language models frequently fail to balance staying truthful with being supportive. They often exhibit sycophancy in responses to users, agreeing with false claims, offering u…
Training LLMs to Verbalize Evaluation Awareness
Usman Anwar, Sahar Abdelnabi, David Krueger
Evaluation awareness (EA) can cause large language models (LLMs) to behave differently during audits than in deployment, yet measuring and accounting for EA remains challenging. We…
SEABench: Benchmarking Endogenous Misalignment In Self-Evolving Agents
Saswat Das, Parvati Viswanathan, Daniel Donnelly +3
Self-evolving LLM agents have gained prominence for their ability to improve after deployment by modifying their harness, including their controller instructions, memory management…
Decomposing and Measuring Evaluation Awareness
Changling Li, Terry Jingchen Zhang, Jie Zhang +4
Frontier language models sometimes recognize that they are under evaluation and adjust their behavior which can undermine validity of benchmark results. Yet the field studies it wi…
Measuring Security Without Fooling Ourselves: Why Benchmarking Agents Is Hard
Sahar Abdelnabi, Chris Hicks, Konrad Rieck +1
The benchmarks used to evaluate AI agents in security-critical roles suffer from crucial weaknesses. Building on recent empirical evidence, we characterize three core challenges th…
Hidden in Memory: Sleeper Memory Poisoning in LLM Agents
Sidharth Pulipaka, Stanislau Hlebik, Leonidas Raghav +4
Large language models are increasingly augmented with persistent memory, allowing assistants to store user-specific information across sessions for personalization and continuity.…
AI Agents May Always Fall for Prompt Injections
Sahar Abdelnabi, Eugene Bagdasarian
Prompt injection is the most critical vulnerability in deployed AI agents. Despite recent progress, we show that the prevailing defense paradigm (data-instruction separation) both…
No More, No Less: Task Alignment in Terminal Agents
Sina Mavali, David Pape, Jonathan Evertz +5
Terminal agents are increasingly capable of executing complex, long-horizon tasks autonomously from a single user prompt. To do so, they must interpret instructions encountered in…
Models That Know How Evaluations Are Designed Score Safer
Katharina Deckenbach, Haritz Puerto, Jonas Geiping +1
The validity of AI safety evaluations depends on models behaving consistently across controlled and deployment settings. Prior work has identified test-time contextual cues, such a…
Skill-Inject: Measuring Agent Vulnerability to Skill File Attacks
David Schmotz, Luca Beurer-Kellner, Sahar Abdelnabi +1
LLM agents are evolving rapidly, powered by code execution, tools, and the recently introduced agent skills feature. Skills allow users to extend LLM applications with specialized…
Stateless Yet Not Forgetful: Implicit Memory as a Hidden Channel in LLMs
Ahmed Salem, Andrew Paverd, Sahar Abdelnabi
Large language models (LLMs) are commonly treated as stateless: once an interaction ends, no information is assumed to persist unless it is explicitly stored and re-supplied. We ch…
Colosseum: Auditing Collusion in Cooperative Multi-Agent Systems
Mason Nakamura, Abhinav Kumar, Saswat Das +5
Multi-agent systems, where LLM agents communicate through free-form language, enable sophisticated coordination for solving complex cooperative tasks. This surfaces a unique safety…
ConVerse: Benchmarking Contextual Safety in Agent-to-Agent Conversations
Amr Gomaa, Ahmed Salem, Sahar Abdelnabi
As language models evolve into autonomous agents that act and communicate on behalf of users, ensuring safety in multi-agent ecosystems becomes a central challenge. Interactions be…
Agent Skills Enable a New Class of Realistic and Trivially Simple Prompt Injections
David Schmotz, Sahar Abdelnabi, Maksym Andriushchenko
Enabling continual learning in LLMs remains a key unresolved research challenge. In a recent announcement, a frontier LLM company made a step towards this by introducing Agent Skil…
Terrarium: Revisiting the Blackboard for Multi-Agent Safety, Privacy, and Security Studies
Mason Nakamura, Abhinav Kumar, Saaduddin Mahmud +3
A multi-agent system (MAS) powered by large language models (LLMs) can automate tedious user tasks such as meeting scheduling that requires inter-agent collaboration. LLMs enable n…
Contextual Integrity in LLMs via Reasoning and Reinforcement Learning
Guangchen Lan, Huseyin A. Inan, Sahar Abdelnabi +5
As the era of autonomous agents making decisions on behalf of users unfolds, ensuring contextual integrity (CI) -- what is the appropriate information to share while carrying out a…
Firewalls to Secure Dynamic LLM Agentic Networks
Sahar Abdelnabi, Amr Gomaa, Eugene Bagdasarian +2
The emergence of agent-to-agent communication protocols mirrors the early internet: powerful connectivity with minimal security infrastructure. When AI agents communicate on behalf…