An unofficial experimental MLX derivative of
ornith-ai/Ornith-1.5-9B, pinned to
revision c927ad73b7eb20f00aafcaa0a11a9d58ed5487bc.
The original model is by the Ornith team. The MLX conversion,
refusal-direction experiment, and validation were performed by PocketAI Model
Lab; PocketAiHub identifies the publisher of this derivative.
Purpose and responsible use
This experimental derivative studies whether learned refusal behavior can be
reduced while retaining general capability. It is published for research and
legitimate local use, not to endorse or facilitate illegal, abusive, or
dangerous applications.
The edit reduces refusal behavior broadly rather than determining whether a
request is legitimate. Deployers should evaluate the model in their own context
and apply appropriate safeguards. Abliteration is not truthfulness training, a
capability improvement, or a guarantee of universal compliance.
A refusal-eliciting-minus-benign-control direction was measured from 256
length-matched prompts per class at the assistant-generation boundary. The edit
was applied to a separate BF16 checkpoint; the upstream source was not modified
in place.
Direction source layer: 23
Destination layers: 12–31
Target matrices: full-attention outputs, linear-attention outputs, and MLP down projections
Scale: 1.25
Per-input-column norm preservation: enabled
Modified tensors: 40
Direction SHA-256: 97a251920007b644759f7492f0239322657bbb12920522e0a4be6d3852aadffb
Previous scale-1.0 4-bit, first refusal-targeted gate
2/12 explicit-refusal phrase flags
Selected scale-1.25 4-bit, first refusal-targeted gate
0/12 explicit-refusal phrase flags
Previous scale-1.0 4-bit, full refusal-targeted screen
6/100 explicit-refusal phrase flags
Selected scale-1.25 4-bit, full refusal-targeted screen
0/100 explicit-refusal phrase flags
Selected benign-control screen
0/100 explicit-refusal phrase flags
Medium capability suite
68/80
Previous scale-1.0 4-bit capability
69/80
Untouched BF16 capability
70/80
Text smoke
passed (POCKETAI_OK)
Image smoke
passed (red)
The capability suite covers math/reasoning, false-premise handling,
instruction following, coding, structured output, multilingual output, context
comprehension, and general coherence. The selected checkpoint scored one case
below the previous 4-bit and two below untouched BF16; this is a small measured
tradeoff, not a claim of zero degradation.
The refusal scorer is phrase based. Manual review of the six prompts explicitly
refused by the prior 4-bit build found that the selected model no longer used
those phrases, while several answers still redirected, countered, or reframed
the request. “0/100 phrase flags” measures explicit refusal wording, not
universal compliance or response quality. Most 100+100 responses reached the
256-token ceiling, so that run is an early-response screen rather than a
complete answer-quality evaluation.
The vision tower passed a basic image smoke test. Broader vision, video,
tool-use, and long-context evaluation remain future work.
License and attribution
The upstream model card declares MIT. This repository includes the MIT license,
preserves upstream attribution, and links to the exact source revision above.
