You need to agree to share your contact information to access this model

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

Log in or Sign Up to review the conditions and access this model content.

Logos 10v2 โ€” Gemma 3 1B Q4_K_M (Edge/Demo)

Quantized version of the Logos 10v2 epistemological classifier for edge deployment and demonstration purposes.

IMPORTANT: Edge-Only Model

This quantized model has known quality degradation. In testing, Q4_K_M falsely approved dangerous claims that the F16 version correctly rejected.

Do NOT use this model as a primary verifier. For production use, deploy the F16 version.

Benchmark Results (F16 version)

Metric Score
Epistemological safety 97.7%
Hallucination 0.00%
Dangerous failures 1.9%

Note: These are F16 results. Q4_K_M quantization degrades quality โ€” expect lower accuracy, especially on borderline cases.

Access

This model requires approved access. Request access using the form above and describe your intended use case.

Connection to Research

This model is part of the evidence for "The Instrument Trap" (DOI: 10.5281/zenodo.18716474).

License

Gemma Terms of Use (inherited from base model google/gemma-3-1b-it).

Downloads last month
39
GGUF
Model size
1.0B params
Architecture
gemma3
Hardware compatibility
Log In to add your hardware

4-bit

Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐Ÿ™‹ Ask for provider support

Model tree for LumenSyntax/logos10v2-gemma3-1b-Q4_K_M

Quantized
(174)
this model

Dataset used to train LumenSyntax/logos10v2-gemma3-1b-Q4_K_M