Research of DeepSeek Engram Architecture based on Qwen-3 and Stable Diffusion series.
Glorious Engrammer keymap for Glove80 keyboard
The original Engram keyboard layout (2021) for English, superseded by Engram-en.
Open-source memory and context for user-aware agents: scoped memory, provenance, retrieval quality, correction, boundaries, evals, and MCP/HTTP access.
Keyboard layout: Engrammer meets Promethium
A smart wallet built on DERO
Dashboard for monitoring and exploring Engram memory: sessions, observations, prompts, and stats
ENGRAFT: write facts into an LLM's n-gram (Engram) memory table as a removable overlay: 100 facts in a 9 MB file, no weights touched, verified on llama.cpp GGUF (Qwen3.8-Flash-Next).
🚀 Engram-PEFT: An unofficial implementation of DeepSeek Engram. Inject high-capacity conditional memory into LLMs via sparse retrieval PEFT without increasing inference FLOPs / DeepSeek Engram 架构的非官方实现。通过参数高效微调 (PEFT) 为大语言模型注入超大规模条件记忆，支持稀疏更新且不增加推理开销。
flashmind — Gemini 3.8 Flash Smart Agent. Gives Gemini Flash persistent memory (Engram), coding skills, and no-hallucination rules. Works in Gemini CLI, Antigravity, Cursor, VS Code, Claude Code and any MCP client.
A user's facts become a few rows in a content-addressed memory table — not a rewrite of the model. Paper, code, data, and interactive site for "User as Engram" (arXiv:2606.19172).
Interactive visualization and exploration of large language model (LLM) systems, infrastructure, and inference workflows
Keyboard-first ODD profiles, model variants, agent suite, session vault and portable task management for OpenCode
Memory for AI harnesses that recalls a multi-graph, not a .md file or RAG chunks and a shrug.
Self-evolving team memory
Rust-native MoE inference runtime with custom CUDA kernels for Blackwell GPUs. Includes DFlash speculative decoding, multi-tier Engram memory, and entropy-adaptive routing. Targets Qwen3.5-35B-A3B on a single RTX 5060 Ti 16GB.
High-performance C++23/Python sparse storage engine for PB-scale recommender system embeddings, Engram, MoE, and KV Cache. Features Hogwild! lock-free updates, slot-directory reads, AVX-512 SIMD, and Lance 2PC checkpointing. Jax, PyTorch are all available.
Multi-agent dev environment: Claude Code thinks, cheaper models type. herdr + OpenCode + OmniRoute + Ollama + Engram.
NGramStore: a general-purpose backend for learned n-gram embeddings — compression, storage and exact lookup. FastText and Engram/TinyEngram are application examples.
The Biggest, Centralized, No BS Open-Source pSEO Content Aggregator for IP University Notes and Resources.
To associate your repository with the engram topic, visit your repo's landing page and select "manage topics."