repoGitHubTrust 82 · PrimaryPublished 15h agoLive · 15h ago
MemTensor/MemOS
Self-evolving memory OS for LLM & AI Agents: ultra-persistent memory, hybrid-retrieval, and cross-task skill reuse, with 35.24% token savings
Lineage graph
Paper → model → repo connections mined from source citations (Tier-1 exact match).
Implements
paperMemSyco-Bench: Benchmarking Sycophancy in Agent MemorypaperSelective Memory Retention for Long-Horizon LLM AgentspaperAlways-OnAgents:A Survey of Persistent Memory, State, and Governance in LLMAgentspaperManufactured Confidence: How Memory Consolidation Turns Hearsay into Confident FactspaperAutoMem: Automated Learning of Memory as a Cognitive Skill
Related across the graph
paperAlways-OnAgents:A Survey of Persistent Memory, State, and Governance in LLMAgentspaperAutoMem: Automated Learning of Memory as a Cognitive SkillpaperManufactured Confidence: How Memory Consolidation Turns Hearsay into Confident FactspaperSelective Memory Retention for Long-Horizon LLM AgentspaperMemSyco-Bench: Benchmarking Sycophancy in Agent Memory