Fermion Research

Research

Research on ternary QAT, packed model formats, local inference, and exact verification across runtimes.

Publications

Technical work

Detta

Announcement

Detta now runs on the Neural Engine

Faster dictation, a sixth of the memory at rest, and the same accuracy.

Read the announcement
A field of warm mineral light gathering into a single bright band

Research

Introducing Phonon-2

The most accurate open speech recognition model under 900 MB, in a 164 MB download that turns an hour of audio into text in about 20 seconds on a MacBook Air.

The Detta window on macOS

Announcement

Detta

A dictation app for the Mac that runs Phonon-2 on the device. Hold a key, speak, and the words appear in any text field.

A field of warm mineral light gathering into a single bright band

Announcement

Introducing Phonon-1

An open speech recognition model for English in a 415 MB download. It runs on a laptop or a datacenter GPU and transcribes an hour of audio in about two and a half minutes.

A cobalt cellular membrane becoming progressively finer

Research

Intelligence at one-eighth the bits

One-shot two-bit conversion lands near chance. Training inside the constraint produces 72.1 MMLU from a 3.88 GB artifact. This is what changed inside the weights.

A honey-toned filament carrying energy across a tactile surface

Research

The Neutrino Engine

An A100 can move the weights for one measured decoder step in 104 microseconds. The step takes 694. The Neutrino Engine is built around the gap.