Memory Agent Benchmarks

Follow this topic Ask about this topic

Momentum

16 papers in the last four weeks, up 167% on the four weeks before. 0.2% of all new papers.

Jul 13Week of Sep 28

Latest papers 89

All topics

Bookkeeping, Composition, or Unreachable Gold? Reading MemoryAgentBench's Conflict-Resolution Scores Against a Frozen Last-Write Resolver
Oct 6, 2026Egor Pakhomov, Erik NijkampMemory Agent Benchmarks
Salesforce AI Research
Persistent Memory in Multi-Agent LLM Inference: What It Costs, What It Buys, and When You Can Tell
Oct 6, 2026Hochan Son, Kyungdoe Han, Jaehan Koh +3LLM Inference Efficiency Multi-Agent LLM Systems
University of California, Los Angeles · University of Wisconsin · HCLTech America
The Right Memory in the Wrong Context: Verifying Retrieval Admissibility in Long-Term Agent Memory
Oct 5, 2026Zi Wang, Xingqiao Wang, Emmanuel Addai +2Memory Agent Benchmarks Long-Term Conversational Memory
University of Arkansas at Little Rock
PACMI: Provenance-Aware Cascading Memory Invalidation for Long-Term LLM Agents
Oct 5, 2026Yiqi Wang, Jiaqi Liu, Jiaqi Zhang +4Data Provenance Memory Agent Benchmarks
University of Southern Queensland · Southern University of Science and Technology · Jiangsu University +2
RealCompanion: Benchmarking Human Understanding from Reasoning over Longitudinal Real-World Conversations
Oct 1, 2026Arman Behnam, Sunglyoung Kim, Jiayi Yu +2Memory Agent Benchmarks User Modeling
Quis Lab, USA · Independent Researcher, China
What Should an Agent Remember? Disentangling Retention from Retrieval in Bounded-Memory Evaluation
Sep 30, 2026Juli HuangAI Agent Evaluation Memory Agent Benchmarks
Stanford University
TAGGRAPH: Tag-Augmented Graphs for Graph Retrieval of Agent Persistent Histories
Sep 29, 2026Yu-Shu Chen, Yu-Jung Liang, Pengtao XieGraph-Based Retrieval Memory Agent Benchmarks
Department of Electrical and Computer Engineering University of California, San Diego La Jolla, CA 92093, USA
Auditable Long-Term Memory: A Deterministic Retrieval Chain Measured at 479/475 of 500 on LongMemEval-S
Sep 29, 2026Christopher J. ChanhnourackLong-Context Retrieval LLM Auditing
Centennial Defense Systems (Archivist)
APM-Bench: Benchmarking Cross-session Persistent Memory for Egocentric Streaming Video Assistants
Sep 29, 2026Jianguo Huang, Jinming Liu, Qiyao Wang +9Streaming Video Understanding Egocentric Video QA
Shanghai Jiao Tong University · Eastern Institute of Technology, Ningbo · Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences +4
Mnemon: Raw Records, Fast Judgments, Slow Thoughts
Sep 28, 2026Guangren WangPersistent Memory for Language Models Memory Agent Benchmarks
When Does Selection Replace Extraction? A Pre-Registered Test of Agent Memory with a Typed Decision Model
Sep 28, 2026Rishabh Sharma, Rishika LallConversational Memory LLM Reranking
Independent Researcher
RoutePrism: Tracing Construction Order Effects in Agent Memory
Sep 28, 2026Dong Xu, Zhangfan Yang, Jiantao Wu +5Agent Memory LLM Agent Memory
School of Artificial Intelligence, Shenzhen University · EasternDawn · School of Computer Science, University of Nottingham Ningbo
TRACE: Governing Memory Validity in Evolving Multi-Agent Systems
Sep 27, 2026Wenjun Xiong, Shengtao Zhang, Shangding Gu +5LLM Agent Memory Persistent Memory for Language Models
Shanghai Jiao Tong University · MemTensor(Shanghai) Technology Co., Ltd. · UC Berkeley +1
CoMemBench: Benchmarking Collaborative Memory Boundaries across Multi-Agent Workflow Topologies
Sep 26, 2026Sen Zhao, Ruiqi Kong, Zuyu Zhang +5Multi-Agent Orchestration Multi-Agent Collaboration
Academy of Advanced Interdisciplinary Studies, Chongqing University of Posts and Telecommunications, Chongqing, China · School of Artificial Intelligence, Chongqing University of Posts and Telecommunications, Chongqing, China · Towngas, China +2
TWIST: A Proposed Benchmark for Intervention Quality in Conversational Memory, with a Human-Validated Draft-Alignment
Sep 23, 2026Subrat PandaConversational Memory Memory Agent Benchmarks
MindTwin
EPGM: Execution Provenance for Budgeted Agent Memory Retrieval
Sep 22, 2026Yiqi Wang, Jinqian Ju, Jiaqi Zhang +4Agent Memory Data Provenance
University of Southern Queensland Australia · Southern University of Science and Technology China · Jiangsu University China +2
DolphinBench: Mapping the Pareto Frontier of Agent Memory
Sep 21, 2026Soumil Rathi, Deshraj Yadav, Taranjeet SinghAgent Memory Persistent Memory for Language Models
Mem0
MemCalib: Benchmarking and Optimizing Memory Use in LLM Agents
Sep 21, 2026Ruike Cao, Fanyu Zhao, Fugen Yao +5LLM Agent Memory Memory Agent Benchmarks
University of Science and Technology of China · Qwen Business Unit of Alibaba · Fudan University
LSREP: A Longitudinal State-Replay Protocol for Evaluating Conversational Memory, with ICE v2 as an Audited Local-First Architecture
Sep 15, 2026Deepesh SonarConversational Memory Memory Agent Benchmarks
Thakur College of Engineering and Technology, Computer Engineering, Mumbai, India
Fortunate Recall: Ontology-Driven Memory Lifecycle Management for Persistent Coherence in LLMs
Sep 9, 2026Ansuman Mullick, Eray TüzünConversational Memory Temporal Reasoning in Language Models
Bilkent University
ICM-Bench: Person-Level Identity Reasoning in Multimodal Agents with Long-Term Memory
Sep 3, 2026Shidu Ren, Yunze Liu, Xing Liu +3Multimodal Memory Memory Agent Benchmarks
University of Toronto · Tsinghua University · Northeastern University +1
When Users Don't Ask: Benchmarking Context-Driven Memory Retrieval in Conversational Agents
Sep 3, 2026Wen-Yu Chang, Yun-Nung ChenConversational Memory LLM Agent Memory
National Taiwan University, Taipei, Taiwan
Making Prospective Memory SLM-Shaped: Typed Intention Stores for Small-Model Agents
Sep 1, 2026Jinqing Zhao, Chengcan WuLLM Agent Memory Small Language Models
Peking University
UTILMEM: Benchmarking Evidence Utilization in Long-Term Conversational Memory
Aug 31, 2026Peijun Qing, Fobo Shi, Soroush VosoughiLLM Agent Memory Memory Agent Benchmarks
Dartmouth College · Wuhan University
Governed Persistent Memory: Source-Bound State Semantics and Fail-Closed Release for Long-Horizon Agents
Aug 12, 2026Guodong XuAgent Memory Long-Horizon Agent Evaluation
Qingdao Guodongxiansheng Network Technology Co., Ltd. (Guˇodòng Xi¯ansheng)
Total Recall at What Cost? Benchmarking the Serving Cost of Agentic Memory Systems
Aug 12, 2026Natchanon Pollertlam, Witchayut KornsuwannawitLLM Serving LLM Agent Memory
Bricks Technology · Thailand
SodaMem: Evidence-Grounded Temporal Graph Memory for LLM Agents
Aug 8, 2026Fengrong Wan, Chengcan Wu, Ningtao LyuMemory Agent Benchmarks Graph-Based Agent Memory
1Peking University
From Test-Time Scaling to Reusable Memory: Measuring Crystallization in Text-to-SQL
Aug 7, 2026Jiaqian Wang, Yutao Qi, Wenjin Hou +2LLM Memory Text-to-SQL
Xidian University · Xidian University Xi’an, Shaanxi, China · Shanghai Jiao Tong University +1
ContextWeave: A Real-World Workflow Benchmark
Aug 5, 2026Bo Wang, Yuqian Yao, Enxi Wang +25Agent Memory Agentic Workflows
1Fudan University · 2Shanghai Innovation Institute · 3ByteDance
Caching for the Future: Scrub Jay Episodic Memory Principles for Agent Memory Systems
Aug 5, 2026Kartikey Singh Bhandari, Aarya Wadhwani, Dhruv Kumar +1Temporal IR Episodic Memory
Birla Institute of Technology and Science, Pilani

Oct 6, 2026, cs.AIJ/K move · Enter open · S save

Bookkeeping, Composition, or Unreachable Gold? Reading MemoryAgentBench's Conflict-Resolution Scores Against a Frozen Last-Write Resolver

Egor Pakhomov, Erik Nijkamp

Salesforce AI Research

Memory Agent Benchmarks

Open paper AskPDF arXiv

MemoryAgentBench's Conflict Resolution split is read as measuring "selective forgetting". We execute the benchmark's own rule - the newest statement about a fact wins - as a zero-learning resolver frozen on one of the four fact lists. Under the official metric the rule answers 80.25% of the questions (74.5% on the three held-out lists). Of the rest, 67 items have a released gold that the last-write graph cannot reach but overwritten statements would ("The capital of India is New Delhi." superseded by "The capital of India is Grosseto."; gold New Delhi); such items are a third of the multi-hop questions at 262K. Two long-context models and our pre-registered approximate re-implementation of the benchmark's BM25 agent, one retained run per item and outcomes only, score 84.7%, 82.6% and 41.6% on the items the rule solves against 10.4%, 11.9% and 6.0% on those 67. The failures are a reachability split plus a small parser-scope residual; the per-item split, not the aggregate, is the unit at which a score here can be read.

Momentum

Related topics

Latest papers 89