Persistent Semantic Entities in Tool-Augmented LLM Systems
Zhaohui Wang
cs.LG, cs.AI, cs.CR, cs.SE
Submitted: 2026-08-08
Updated: 2026-08-11
Comments: Accepted at the 43rd International Conference on Machine Learning (ICML 2026). Camera-ready version; 32 pages . The openreview url is https://openreview.net/forum?id=zPjmtawzT8¬eId=CXB95ws5so and ICML poster url is https://icml.cc/virtual/2026/poster/60521
Code: https://github.com/GeoffreyWang1117/PSE-ICML2026
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- Memory Poisoning Attack and Defense on Memory Based LLM-Agents
- Memory Injection Attacks on LLM Agents via Query-Only Interaction
- Constitutional AI: Harmlessness from AI Feedback
- MemGPT: Towards LLMs as Operating Systems
- MindGuard: Intrinsic Decision Inspection for Securing LLM Agents Against Metadata Poisoning
- MCPTox: A Benchmark for Tool Poisoning Attack on Real-World MCP Servers
- AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation
- InjecAgent: Benchmarking Indirect Prompt Injections in Tool-Integrated Large Language Model Agents
- AgentSentry: Mitigating Indirect Prompt Injection in LLM Agents via Temporal Causal Diagnostics and Context Purification
- The Instruction Hierarchy: Training LLMs to Prioritize Privileged Instructions
- PoisonedRAG: Knowledge Corruption Attacks to Retrieval-Augmented Generation of Large Language Models
Related papers
- Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
- AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
- Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
- Convergence issues in Relational Concept Analysis based on AOC-posets
- Beliefs Beyond Posteriors: The Predictive Cost of ELBO Binding
- Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks