E
EvoKernel
Self-evolving framework for NPU kernel synthesis without fine-tuning or large datasets
Open SourceFree
About
EvoKernel is a research framework that generates optimized kernels for Domain-Specific Architectures like NPUs using a self-evolving agentic approach. Unlike traditional code generation models requiring extensive fine-tuning, it employs value-driven memory to overcome cold-start problems in data-scarce environments. Particularly valuable for hardware teams working with specialized accelerators where training data is limited.
Details
| Type | coding-agent |
| Deployment | self-hosted |
| Supported Models |
Tags
autonomouscoding-agentmemoryself-hostedopen-sourceresearch
Quick Info
- Organization
- Research
- Pricing
- open-source
- Free Tier
- Yes
- Updated
- Mar 12, 2026
Also in Agents
A
AI Data Analysis Agent
Autonomous agent that analyzes datasets and generates visual insights
OSSFree
Shubham Saboo
104.2Ktoday74
A
AI Deep Research Agent
Autonomous agent that conducts comprehensive multi-source research investigations
OSSFree
Shubham Saboo
104.2Ktoday74
A
AI Journalist Agent
Autonomous agent that researches topics and writes structured news articles
OSSFree
Shubham Saboo
104.2Ktoday74