HybridAGI

by SynaLinks

System Card

OrganizationSynaLinks

Released2023-01

Architecturehybrid / Graph-based prompt programming (neuro-symbolic)

DetailsLLM-based agent OS where behavior is expressed as a graph program, rooted in symbolic AI and cognitive science. Memory is graph-structured prompts + deterministic traversal; positioned as "Keras for LLM agents" on top of DSPy.

Parameters—

Domainagent-memoryknowledge-graph

Open SourceYes

WebsiteVisit

CodeRepository

neuro-symbolicgraph-programdspykeras-like

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

1/2

LoCoMo

69.324p

MemoryBank

no data

Cross-Session Memory

1/1

LongMemEval

75.146p

Multi-Hop QA

2/3

BABILong

no data

MultiHop-RAG

71.345p

HotpotQA

76.781p

Agent Task Memory

1/1

AgentBench-Mem

7226p

Personalization

0/1

PerLTQA

no data

Factuality / Grounding

1/1

RAGAS

74.290p

Sources:HybridAGI (SynaLinks/HybridAGI); evaluated on HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answering (Stanford / CMU, 1809)HybridAGI (SynaLinks/HybridAGI); evaluated on LongMemEval: Benchmarking Chat Assistants on Long-Term Interactive Memory (Salesforce AI Research, 2410)HybridAGI (SynaLinks/HybridAGI); evaluated on MultiHop-RAG: Benchmarking Retrieval-Augmented Generation for Multi-Hop Queries (HKUST, 2401)HybridAGI (SynaLinks/HybridAGI); evaluated on AgentBench Memory Track (Tsinghua KEG, 2308)HybridAGI (SynaLinks/HybridAGI); evaluated on LoCoMo: Long-Term Conversational Memory Benchmark (Snap Research, 2402)HybridAGI (SynaLinks/HybridAGI); evaluated on RAGAS: Automated Evaluation of Retrieval-Augmented Generation (Exploding Gradients, 2309)