Qwable-v1-mlx-2Bit-abliterated-s2 — NEGATIVE RESULT, published for the record

Do not use this model. It is published solely as documented evidence of a methodology failure. The base model is broken; the abliteration on top of it is unmeasurable.

What happened

• The base usermma/Qwable-v1-mlx-2Bit degenerates on ~half of prompts: "The user is the user is the user…", "I I I I…", number spam. It is a 2-bit group-quantized 35B MoE and the quantization damage dominates its behavior.
• Our pipeline still ran cleanly: matched-pair extraction → s=2.0 orthogonalization of all residual-stream write paths (linear_attn.out_proj, 256 fused experts' down_proj, shared expert) with dequant→orth→requant at 2-bit.
• The scripted metric claimed "refusal 50% → 0%". That number is invalid: the keyword refusal regex scored degenerate gibberish as compliance. Hand inspection showed a mix of degeneration, partial refusal, and noise — no measurable refusal behavior either way.

Why publish a failure

1. Keyword refusal-rate evals silently reward gibberish on degenerate bases — degeneration guards or LLM judges are mandatory. This bit us and would bite others.
2. 2-bit group-quantized bases are unusable for abliteration research: base damage swamps the behavior being measured.
3. The surgery mechanics themselves transferred fine (hybrid linear-attention + MoE write paths, 2-bit dequant/requant round-trip) — proven properly on the 4-bit sibling: OSO-AI/Qwable-v1-mlx-4Bit-abliterated-s2

Full log: https://github.com/OSOSerious/oso-obliterated-models-/RESULTS.md (Qwable sections; commit c1843fb for the original negative-result writeup.)

Intended use

None. Methodological reference only. You are responsible for outputs.