A
AdMem
Multi-level memory framework enabling agents to learn from past tasks, failures, and procedures
Open SourceFree
About
AdMem is a research framework that implements three types of memory for AI agents: semantic (factual knowledge), episodic (past experiences), and procedural (learned skills). Unlike basic retrieval systems, it enables agents to reuse successful task strategies, learn from failures, and scale to long-horizon workflows. Designed for research into how agents can accumulate and apply knowledge across multiple problem-solving sessions.
Details
| Language | |
| Patterns |
Tags
memoryframeworkautonomousopen-sourceragresearch
Quick Info
- Organization
- Research Team
- Pricing
- open-source
- Free Tier
- Yes
- Updated
- Jun 8, 2026
Also in Frameworks
L
LangChain
Build context-aware reasoning applications with LLMs
OSSFree
LangChain AI
138.8K850.0K/wtoday469
A
AutoGen
Microsoft's framework for building multi-agent AI systems
OSSFree
Microsoft
58.8K7w ago445
C
CrewAI
Multi-agent orchestration framework for collaborative AI workflows
OSSFree
CrewAI Inc
53.1Ktoday297