Back to Arena

MemoryBank

by Institute of Software, Chinese Academy of Sciences

System Card

OrganizationInstitute of Software, Chinese Academy of Sciences
Released2023-05
Architecturevector-rag / Ebbinghaus-curve forgetting mechanism
DetailsMemory store with retrieval and update mechanisms inspired by the Ebbinghaus forgetting curve — permits the AI to forget and reinforce memories based on elapsed time and significance. Summarizes past events and synthesizes user personality.
Parameters—
Domainagent-memorypersonalizationlifelong-learning
Open SourceYes
PaperView Paper
CodeRepository
ebbinghausforgetting-curvesiliconfriendcompanionaaai-2024

Capability Profile

Benchmark Scores

6 of 14 benchmarks
Long-Context Retrieval
0/5
RULER
no data
NIAH
no data
LooGLE
no data
LongBench
no data
∞Bench
no data
Multi-Turn Recall
2/2
LoCoMo
62.314p
MemoryBank
81.888p
Cross-Session Memory
1/1
LongMemEval
60.113p
Multi-Hop QA
1/3
BABILong
59.73p
MultiHop-RAG
no data
HotpotQA
no data
Agent Task Memory
1/1
AgentBench-Mem
66.815p
Personalization
1/1
PerLTQA
72.623p
Factuality / Grounding
0/1
RAGAS
no data
Sources:arXiv:2305.10250 Table 2 — SiliconFriend-ChatGPT English Ranking 0.818 rescaled to 0-100MemoryBank paper (arXiv:2305.10250); evaluated on LoCoMo: Long-Term Conversational Memory Benchmark (Snap Research, 2402)MemoryBank paper (arXiv:2305.10250); evaluated on LongMemEval: Benchmarking Chat Assistants on Long-Term Interactive Memory (Salesforce AI Research, 2410)MemoryBank paper (arXiv:2305.10250); evaluated on PerLTQA: A Personal Long-Term Memory Question Answering Dataset (PolyU, 2402)MemoryBank paper (arXiv:2305.10250); evaluated on AgentBench Memory Track (Tsinghua KEG, 2308)MemoryBank paper (arXiv:2305.10250); evaluated on BABILong: Testing the Limits of LLMs with Long-Context Reasoning-in-a-Haystack (AIRI, 2406)