AI

An LLM inside an LLM - Same guy who put a RISK-V CPU inside Qwen has now made Karpathy's stories260k run on an NPU compiled inside Qwen 3.5-4B's weights

Read full article at source ↗

Based on a feed excerpt; full article content has not been acquired. Source evidence retrieved 31/10/2026, 22:47:25 UTC.

Briefings

A user named LatentSpacer created a QwenPU model running Karpathy's stories260k on an NPU compiled into Qwen 3.5-4B weights. The model uses Qwen's internal memory for weights and performs bit-exact matmuls, softmax, and attention operations. The project is still in development with ongoing optimizations and size reductions.

Key points

• A user named LatentSpacer created a QwenPU model running Karpathy's stories260k on an NPU compiled into Qwen 3.5-4B weights.
• The model uses Qwen's internal memory for weights and performs bit-exact matmuls, softmax, and attention operations.

Context

This project demonstrates an LLM running inside an LLM, specifically running Karpathy's stories260k on an NPU compiled into Qwen 3.5-4B weights. The model uses Qwen's internal memory for weights and performs bit-exact matmuls, softmax, and attention operations.

Limitations

The evidence is a single tweet from October 2026, with the full writeup still a work in progress. The project is in development with ongoing optimizations and size reductions, and specific performance metrics or deployment details are not provided.

Generated by Bittrees-hosted qwen35-2b sha256:aaf42c8b7c3c on 01/11/2026, 00:23:04 UTC. Based on publisher-provided feed evidence; may not cover the full original article. Check the source for details.

View archived briefing on IPFS
bafybeifdfkfowdzfocj3dmuhej6znof6ijdmhxxjgnptcnor63ro6ghkbe