8 papers
When Is a Memory Safe Enough to Act On? Action-Conditioned Verification for Long-Lived Agents
Yiqi Wang, Yifei Wang, Jiaqi Zhang +4
Long-lived agents act on memory carried across sessions, and many of their actions are irreversible, so a stale memory can turn directly into harm. Existing work improves the memor…
Learning What to Ask for Tomorrow: Active Preference Acquisition for Long-Term Personalized Web Agents
Yiqi Wang, Zizhi Zeng, Jiaqi Zhang +3
Web agents increasingly serve the same user across repeated online tasks such as booking travel, shopping, and scheduling, so their long-term usefulness depends on how well they kn…
PACMI: Provenance-Aware Cascading Memory Invalidation for Long-Term LLM Agents
Yiqi Wang, Jiaqi Liu, Jiaqi Zhang +4
LLM agents rely on long-term memory to retain and reuse information when performing tasks over long horizons. Existing methods provide limited support for handling memories that be…
When Does Execution Provenance Help Agent Memory Retrieval?
Yiqi Wang, Jinqian Ju, Jiaqi Zhang +4
A language agent's execution history can exceed its context window, requiring its memory system to retrieve complete supporting evidence under a hard token budget. Evidence may spa…
MAP-Graph: Provenance-Aware Shared Memory for Multi-Agent Workflows
Yiqi Wang, Zihao Yan, Jiaqi Zhang +5
Shared memory helps language-model agents reuse information across long workflows, yet relevant evidence may not be admissible for a particular agent or action. Because restriction…
From Faulty Memories to Corrected Actions: Dependency-Guided Rollback Repair for Memory-Augmented Agents
Caili Yu, Yiqi Wang, Jiaqi Zhang +5
Persistent memory lets language model agents reuse information across sessions, but it also makes errors durable: a poisoned, stale, or misattributed record can alter reasoning, to…
COVENANT: Natural-Language Workflow Compilation for Aligned Agent Execution
Jincheng Wang, Min Zheng, Tao Wei
Large language model (LLM) agents are increasingly entrusted with natural-language workflow instructions (e.g., retail-payment policies) that specify not only what outcome to achie…
MemTX: Transactional Belief Commit for Stateful Agent Memory
Xiaoyang Li, Yiqi Wang, Haohui Lu +5
LLM agents increasingly coordinate through persistent shared memory: one agent's write becomes another agent's premise, and eventually a tool call with real side effects. Current a…