arXiv Computation and Language

LLMs Can't Play Hangman: On the Necessity of a Private Working Memory for Language Agents

The paper argues that large language models (LLMs) lack a private working memory, which is essential for tasks that require hidden state, such as interactive games. It introduces Private State Interactive Tasks (PSITs) and proves that chat-based LLMs cannot maintain a secret state while producing consistent public responses. Empirical tests confirm that standard LLMs and retrieval-based baselines fail, while a new architecture with explicit private memory succeeds.

arXiv AI
2d ago

PrivDrift: Auditing User-Secret Leakage Under Topic Drift in Active LLM Conversations

PrivDrift is a benchmark that tests whether user‑disclosed secrets can still be recovered by large language models after the conversation shifts to unrelated topics. It includes 1,000 controlled multi‑turn dialogues with seeded secrets, topic‑drift turns, and standardized extraction probes. Experiments on three LLMs with extended context windows show that dialogue‑level leakage remains substantial—between 38.7% and 54.6%—and is influenced by model, secret type, and persuasion intensity, while additional topic drift does not reliably reduce leakage.

By Luciano Maldonado
arXiv Machine Learning
Sep 14

AIM: A Privacy-Aware Interoperable Memory Framework for Multi-Agent Multi-User LLM Systems

The paper introduces AIM, a privacy‑aware memory framework that lets multi‑agent, multi‑user large language models manage both private and shared memory. AIM classifies data as private (user‑specific) or public (shared) and enforces index‑level access controls to protect sensitive information while enabling shared knowledge to improve coordination. The authors also present MUMBench, a new dataset for evaluating memory operations in multi‑user settings, and report high accuracy metrics for AIM on this benchmark.

By Zachary Johnson, Nigel Boachie Kumankumah, Somya Chatterjee, Tejas Sathyamurthi, Min Chen, Xinyi Alice Li, Xiao Wang, Emily Morgan Gelchie, Jessica Lin, Sadid A. Hasan, Sulaiman Vesal
arXiv AI
Aug 28

SKILL.state: Scalable Long-Horizon Agent Skills

SKILL.state is a new runtime architecture for large language model agents that replaces the traditional append‑only conversational history with an explicit, mutable execution state. At each step the model receives only the immutable skill specification, the current structured state, and the latest observation, discarding intermediate reasoning after validating state updates. Experiments across datasets, models, and environments show that SKILL.state improves task accuracy and significantly reduces cumulative token consumption, proving that explicit execution state is a scalable, architecture‑agnostic abstraction for long‑horizon agent skills.

By Sanket Badhe, Priyanka Tiwari, Jonghyun Chung
Hugging Face Trending Papers
Aug 3

MNC: Scope-Bound Semantic Declassification for Private LLM-Agent Communication

Multi-agent large language model (LLM) systems can expose protected state through internal messages, tool arguments, logs, and persistent memory even when their public outputs appear innocuous. Existing privacy prompts, redaction methods, and source-level access controls restrict surface content or data access, but do not specify what a legitimately informed agent should disclose or how that disclosure may be reused downstream.

arXiv AI
Sep 3

Architecting Conversational Data Systems for Stateless LLM APIs: The Hydration Proxy Pattern

The paper introduces the Hydration Proxy Pattern, an architectural approach that separates session persistence from the reasoning engine in conversational AI systems. This pattern allows enterprise platforms to maintain sovereignty over conversational data while still leveraging stateless LLM APIs for scalability. Additionally, the authors propose the Context Stabilization Mandate to balance sovereign state management with key‑value caching.

By Joseph Axisa
arXiv Computation and Language
6d ago

MemoryArena: Benchmarking Agent Memory in Interdependent Multi-Session Agentic Tasks

MemoryArena is a new evaluation gym that benchmarks agent memory in interdependent multi‑session tasks. Unlike prior benchmarks that test memorization or single‑session action in isolation, MemoryArena requires agents to acquire memory while interacting with the environment and then use that memory to guide future decisions across a range of tasks such as web navigation, planning, information search, and formal reasoning. The benchmark reveals that agents excelling on existing long‑context memory tests perform poorly here, highlighting a gap in current memory evaluation methods.

By Zexue He, Yu Wang, Churan Zhi, Yuanzhe Hu, Tzu-Ping Chen, Lang Yin, Ze Chen, Tong Arthur Wu, Siru Ouyang, Zihan Wang, Jiaxin Pei, Julian McAuley, Yejin Choi, Alex Pentland