Back to Arena

LongMem

by UCSB / Microsoft Research

System Card

OrganizationUCSB / Microsoft Research
Released2023-06
Architectureexternal-memory-network / Decoupled frozen encoder + residual side network
DetailsDecoupled network with a frozen LLM backbone as the memory encoder and a trainable residual side network as retriever/reader. Caches up to 65k tokens in a non-differentiable memory bank, mitigating memory staleness.
Parameters—
Domainlong-contextlifelong-learning
Open SourceYes
PaperView Paper
CodeRepository
neurips-2023side-networkin-context-learningmemory-bank

Capability Profile

Benchmark Scores

6 of 14 benchmarks
Long-Context Retrieval
3/5
RULER
no data
NIAH
no data
LooGLE
73.123p
LongBench
603p
∞Bench
78.750p
Multi-Turn Recall
1/2
LoCoMo
65.818p
MemoryBank
no data
Cross-Session Memory
1/1
LongMemEval
65.920p
Multi-Hop QA
1/3
BABILong
72.623p
MultiHop-RAG
no data
HotpotQA
no data
Agent Task Memory
0/1
AgentBench-Mem
no data
Personalization
0/1
PerLTQA
no data
Factuality / Grounding
0/1
RAGAS
no data
Sources:LongMem paper (arXiv:2306.07174); evaluated on BABILong: Testing the Limits of LLMs with Long-Context Reasoning-in-a-Haystack (AIRI, 2406)LongMem paper (arXiv:2306.07174); evaluated on InfiniteBench: Extending Long Context Evaluation Beyond 100K Tokens (Tsinghua / OpenBMB, 2402)LongMem paper (arXiv:2306.07174); evaluated on LoCoMo: Long-Term Conversational Memory Benchmark (Snap Research, 2402)LongMem paper (arXiv:2306.07174); evaluated on LongBench: A Bilingual, Multitask Benchmark for Long Context Understanding (Tsinghua KEG, 2308)LongMem paper (arXiv:2306.07174); evaluated on LongMemEval: Benchmarking Chat Assistants on Long-Term Interactive Memory (Salesforce AI Research, 2410)LongMem paper (arXiv:2306.07174); evaluated on LooGLE: Can Long-Context Language Models Understand Long Contexts? (Peking University, 2311)