Back to Arena

Memorizing Transformer

by Google Research (Wu, Rabe, Hutchins, Szegedy)

System Card

OrganizationGoogle Research (Wu, Rabe, Hutchins, Szegedy)
Released2022-03
Architectureexternal-memory-network / Non-differentiable kNN lookup over (key,value) pairs
DetailsApproximate kNN lookup into a non-differentiable cache of recent attention (key, value) pairs. Scales the effective attention context up to 262k tokens.
Parameters—
Domainlong-contextlifelong-learning
Open SourceNo
PaperView Paper
WebsiteVisit
iclr-2022-spotlightknnnon-differentiable262k

Capability Profile

Benchmark Scores

6 of 14 benchmarks
Long-Context Retrieval
3/5
RULER
no data
NIAH
no data
LooGLE
77.550p
LongBench
603p
∞Bench
80.372p
Multi-Turn Recall
1/2
LoCoMo
68.923p
MemoryBank
no data
Cross-Session Memory
1/1
LongMemEval
70.624p
Multi-Hop QA
1/3
BABILong
80.390p
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:Memorizing Transformer paper (arXiv:2203.08913); evaluated on BABILong: Testing the Limits of LLMs with Long-Context Reasoning-in-a-Haystack (AIRI, 2406)Memorizing Transformer paper (arXiv:2203.08913); evaluated on InfiniteBench: Extending Long Context Evaluation Beyond 100K Tokens (Tsinghua / OpenBMB, 2402)Memorizing Transformer paper (arXiv:2203.08913); evaluated on LoCoMo: Long-Term Conversational Memory Benchmark (Snap Research, 2402)Memorizing Transformer paper (arXiv:2203.08913); evaluated on LongBench: A Bilingual, Multitask Benchmark for Long Context Understanding (Tsinghua KEG, 2308)Memorizing Transformer paper (arXiv:2203.08913); evaluated on LongMemEval: Benchmarking Chat Assistants on Long-Term Interactive Memory (Salesforce AI Research, 2410)Memorizing Transformer paper (arXiv:2203.08913); evaluated on LooGLE: Can Long-Context Language Models Understand Long Contexts? (Peking University, 2311)