moenupa/sdpo-cl
Dense token-level supervision accelerates specialization, but can hurt continual learning.
01 / MACHINE INTELLIGENCE
Models you can run. Agents you can inspect. Tools you can build on.
← All topicsDense token-level supervision accelerates specialization, but can hurt continual learning.
[KDD 2026] MemGraphRAG: Memory-based Multi-Agent System for Graph Retrieval-Augmented Generation
Weights that never freeze. A continual-learning architecture, and the pre-registered experiment that falsified its central claim. https://arxiv.org/abs/2607.20792
Hybrid RAG system combining vector search, knowledge graph (LightRAG), and cross-encoder reranking — with Docling document parsing, visual intelligence (image/table captioning), agentic streaming chat, and inline citations. Powered by Gemini or local Ollama models.
Your agent reads the rules. This checks whether it followed them.
One macOS app for Claude Code, Codex, and every agent runtime you use — scheduled runs, global hotkey launcher, per-run git worktrees, one review board.
一个可以让像deepseek-v4这样的纯文本模型, 在codex中无障碍调用内置看图工具(view_image)的方案,附为纯文本llm设计的视力工具包。
Learn what AI writes, not just accept it. A Claude Code skill that turns AI-generated code into educational deep dives.
Windows 本地多模态 AI 桌面桌宠,支持屏幕与音频感知。.
Claude Code plugin that orchestrates GPT Codex agents for implementation, monitoring, and independent reviews.
AI work that doesn't evaporate. A workspace for building, tuning, and publishing governed AI products.
PyTorch Code for Explorative Modeling: Unlocking a Third Pretraining Axis and End-to-End Generation
camelAI — an AI coding assistant platform built on Cloudflare Workers and Durable Objects
Story Creation AI Agent for novel, scripts, translation, interactive games, and IP content
한국인을 위한 스킬 모음집 - 에이전트를 한국인으로.
code for "Tying the Loop - Tied Expert Layers in Mixture-of-Experts Language Models"
Open the project to explore the idea and the code.
Official Repo for "AgenticASR: Refining Speech Recognition in Real-World Scenarios via an Agentic Approach"
Weak-to-Strong On-Policy Distillation.
Run Kimi K3 (2.8T MoE) on a 128GB-RAM consumer machine in pure Rust, experts streamed from disk. Tiny engine, immense model.
Your ultimate agentic markdown workspace.
面向 AI 影视创作的开源无限画布工作台,集成多模态生成、分镜编排、素材管理与 Agent 工作流。
AI agent autonomously optimizing RAG retrieval pipelines against your own documents
Executable, measurable, and reproducible AI4AI toward recursive self-improvement. Home of OpenMLE and Frontis-MA1.