Showcase / question: a board-proven offline language runtime on ESP32-C3, and whether some language capability may eventually move beyond low-bit dense inference

[Alpha-Guardian]

Hi BitNet folks,

I wanted to share a small but unusual language-runtime project that may still be relevant to the broader question of how far language capability can be compressed and deployed, even though it does not follow the low-bit dense model path.

We built a public demo line called Engram and deployed it on a commodity ESP32-C3.

Current public numbers:

• Host-side benchmark capability

  • LogiQA = 0.392523
  • IFEval = 0.780037

• Published board proof

  • LogiQA 642 = 249 / 642 = 0.3878504672897196
  • host_full_match = 642 / 642
  • runtime artifact size = 1,380,771 bytes

Important scope note:

This is not presented as unrestricted open-input native LLM generation on MCU.

The board-side path is closer to a flash-resident, table-driven runtime with:

• packed token weights
• hashed lookup structures
• fixed compiled probe batches
• streaming fold / checksum style execution over precompiled structures

So this is not a standard low-bit dense inference stack. It is closer to a task-specialized language runtime whose behavior has been crystallized into a compact executable form under severe physical constraints.

Repo:
https://github.com/Alpha-Guardian/Engram

Why I’m posting here is that BitNet seems to represent one of the clearest public efforts around pushing language-model efficiency to a much more radical point while still staying within the dense model family.

What I’d be curious about is whether systems like this should be thought of as:

• completely outside the low-bit LLM family
• a different endpoint of capability compression, where some task capability is better deployed as a dedicated executable form rather than an increasingly efficient dense model
• or an early sign that future language systems may combine both low-bit dense models and highly specialized runtimes for certain capability slices

If this direction is relevant to your team, I’d be glad to compare notes.