xiaomi-research/spatio-lm
[ICML 2026 Oral] SpatioLM: Towards General Physical Spatial Intelligence in Vision-Language Models
01 / MACHINE INTELLIGENCE
Models you can run. Agents you can inspect. Tools you can build on.
← All topics[ICML 2026 Oral] SpatioLM: Towards General Physical Spatial Intelligence in Vision-Language Models
Official code for ReBA:Better load balancing for VLM-MoE
Autonomous GUI agent — give it a task, it operates the desktop. Visual memory, one-shot UI learning. | 自主GUI代理——给它一个任务,它操作桌面。视觉记忆,一次学习即可操作。
Official repository of UniPASE, a SOTA USE model.
Open the project to explore the idea and the code.
A Benchmark for Evaluating Agent Self-Evolution on Real Business Tasks
[ICML 2026] Multimodal deep-research MLLM and benchmark. The first long-horizon multimodal deep-research MLLM, extending the number of reasoning turns to dozens and the number of search-engine interactions to hundreds.
PhyAI is a high-performance framework for running Physical AI models (VLA, WAM, and beyond), supporting both cloud-based serving and on-device deployment.
100+ curated Seedance 2.5 prompts with real video previews, plus an installable Agent Skill that optimizes prompts, creates storyboards, and generates videos via Seedance models.
Open the project to explore the idea and the code.
Fine-tune LLMs from one YAML. Layer streaming trains an 8B model on a 4 GB laptop GPU.
Swift + Metal MoE inference for Apple Silicon: Gemma 4 26B in ~2 GB, Qwen 3.6 35B in ~1.45 GB, and DeepSeek-V4-Flash 284B at up to 4.8 tok/s on a 24 GB Mac. SSD-streamed experts, native Mac app, CLI, and OpenAI-compatible server.
Forge is an open-source, self-hosted alternative for visually building and shipping AI agents and workflows without giving up control of your infrastructure.
Spectrum-based acceleration for ComfyUI’s native MiniMax H3 audio-video model. Forecasts post-transformer features with Chebyshev ridge regression to skip selected transformer evaluations, with adaptive scheduling, sampler-aware safeguards, CPU/VRAM history storage, and native fallbacks.
CodeGraph builds a semantic graph of your codebase — functions, classes, imports, call chains — and exposes it through 42 MCP tools, 38 languages, a VS Code extension, and a persistent memory layer. AI agents get structured code understanding instead of grepping through files.
Official Implementation of VisualClaw: A Real-Time, Personalized Agent for the Physical World
This is a ComfyUI plug-in for MinMaxH3.
Open the project to explore the idea and the code.
SmartGallery DAM is a local-first, browser-based Digital Asset Manager for ComfyUI. Generate without opening ComfyUI. Query your gallery with AI, in natural language. Advanced file manager, filter by prompt, model, LoRA, date and comment. Status tags, ratings, virtual collections, client sharing. Mobile-friendly, cross-platform, Docker-ready.
Novel Engine is a desktop app that lets you construct novels. Seven AI agents handle the editorial pipeline — from pitch to published manuscript — while you focus on the story. Powered by Claude, Codex, and Ollama. Open source. Free. No gatekeeping.
NV Broadcast — Unofficial NVIDIA Broadcast for Linux and other OS. AI-powered virtual camera with background removal, blur, replacement, video enhancement, and noise cancellation. GPU accelerated. Open source.
A curated collection of machine learning and AI lecture notes from the world's leading universities. This repository gives you access to the same lecture notes used by students at top institutions such as MIT, helping you learn from the very best educational resources available.
repo for QuerySplat.
Lightweight loop engineering state kernel for long-running AI agent teams. Agent-loop agnostic across Codex, Claude Code, and other coding agents, with durable goals, quota-aware auto-wake, executable todos, evidence logs, and verifiable handoffs.