slvDev
Featured
esp32-ai
A 28.9 million parameter language model runs on an $8 ESP32-S3 microcontroller entirely on-device, generating simple stories at about 9.5 tokens per second.
MatrAIx-Persona-8B
MatrAIx is a population-scale, persona-driven evaluation infrastructure that instantiates heterogeneous simulated users as LLM agents to test AI systems and interactive products across Survey, AI Chatbot, Web, and App environments.
Stars
21
7-day growth
No data
Forks
0
Open issues
0
License
Apache-2.0
Last updated
2026-08-01
AI repository intelligence
FR-AI / ANALYSISWhy it is worth attention
It combines a 1,290-dimensional persona schema with a quality-filtered, publicly released one-million-persona coreset, and offers reproducible task execution, shared telemetry, and a visual Playground.
Who it is for
- AI evaluation and safety researchers
- Product and UX teams validating interactive experiences
- LLM agent developers needing persona-conditioned test scenarios
- Academic communities studying simulated-user methodologies
Use cases
- Running reproducible persona-driven evaluations across survey, chatbot, web, and native app tasks
- Stress-testing AI products with heterogeneous simulated users before real-world release
- Using the Persona 1M dataset for research on synthetic persona generation and grounding
- Developing and registering custom evaluation tasks with verifiers for Playground/CLI runs
Strengths
- Detailed 1,290-dimensional persona schema covering background, psychology, capability, and behavior
- One-million-persona public coreset with dependency-aware synthetic generation and evidence-aware human grounding
- Four built-in task environments plus adapters and task-owned verification
- Apache-2.0 licensed with Docker/uv-based reproducible Harbor job recipes and visual Playground
Considerations
- Explicitly positioned as a simulation tool, not a replacement for evidence from real people
- Requires a multi-tool setup (Docker, uv, Python 3.12, Node.js) and model API keys for persona-agent examples
- Large generated datasets are not in git and must be downloaded separately; custom task development involves multiple components
README quick start
Installation
Description
Simulate Before Reality.
Related repositories
Similar projects matched by category, topics, and programming language.
gavamedia
Featured
deltafin
Deltafin is a research project that runs the 2.8-trillion-parameter Mixture-of-Experts model Kimi K3 on a single Apple Silicon Mac (e.g., M1 Max with 64 GB) at about 16 seconds per token, using exact, reproducible inference with local or streaming expert loading.
jamesob
Featured
local-llm
A comprehensive guide for building and configuring a high-end local machine to run state-of-the-art LLMs, with detailed hardware choices, BIOS tuning, and Docker-based model serving.