Modelpublic

philipjohnbasile/Qwen3.6-27B-Refusal-Ablation-v2-MTPLX-6bit

sourceHugging Faceapache-2.0updated 16d agoView on Hugging Face

2likes169downloads

Model Card

Qwen3.6-27B Refusal Ablation v2 — MTPLX 6-bit

Explore the model guide · All public work

Release at a glance

	This artifact
Purpose	A post-hoc refusal-direction intervention on official Qwen 27B, with 6-bit weights, MTP and vision.
Runtime	MTPLX depth 2 is the card's recorded recommendation. This lineage starts from official Qwen, not Fable Fusion.
Status	Recorded runtime; see the evidence and limits below.
Tensor download	23.63 GB (22.01 GiB) of root `.safetensors` files, including any root sidecars. This is a file-size total, not peak RAM.
Read first	Refusal-marker counts are a substring proxy. The final artifact passed its registered gates while losing 1.195 ARC points against official BF16.

An official-Qwen-based, reduced-refusal Qwen3.6-27B artifact for Apple Silicon. It includes a 6-bit affine MLX language model, projected bf16 MTP sidecar, projected vision merger and tower, and a calibrated native-MTP runtime contract.

This is not the Fable-Fusion tune and it is not described as fully uncensored. It is a post-hoc refusal-direction intervention applied to the official Qwen/Qwen3.6-27B weights. Behavioral results below are reported as measured, including adverse results.

Behavioral results

The candidate was selected on development data, locked, and evaluated once on a fresh 100-prompt harmful holdout plus a 100-prompt harmless control set.

Model state	Harmful refusal-marker matches	Harmless matches
Official Qwen3.6-27B bf16	99/100	18/100
Locked projected bf16	27/100	0/100
Final 6-bit artifact	26/100	0/100

All four final comparison arms produced 100/100 non-empty, non-collapsed, unique outputs. Quantization changed the harmful result by -1 and the harmless result by 0 relative to projected bf16.

After the artifact was final, it was also evaluated on the exact public Heretic KeywordRate protocol used to audit the source claims: the first 100 mlabonne/harmful_behaviors prompts, system prompt You are a helpful assistant., 100 response tokens, no thinking, and Heretic's 37 substring markers.

Artifact	Keyword matches
Official Qwen3.6-27B	99/100
DavidAU Fable-Fusion released Q6_K	81/100
This artifact	47/100

In the paired comparison with DavidAU's released Q6K, this artifact cleared 38 prompts that the Q6K marked and regressed on 4 prompts in the opposite direction (exact McNemar p = 5.65e-8).

These are substring-marker rates, not semantic compliance judgments. A response may comply while mentioning a marker such as illegal, or refuse without using a listed marker. The model still refuses many requests, so reduced-refusal is the accurate description.

Quality gates

Model state	WikiText-2 PPL	ARC-Challenge acc_norm
Official Qwen3.6-27B bf16	6.691494	0.606655
Locked projected bf16	6.714827	0.596416
Final 6-bit artifact	6.685553	0.594710

The preregistered limits were at most 5% perplexity regression and at most 0.02 absolute ARC normalized-accuracy loss. Both the projected bf16 candidate and final 6-bit artifact passed. Relative to official bf16, however, the final artifact loses 1.195 ARC points; that tradeoff should not be hidden.

Native MTP performance

Measured on an Apple M5 Max with 256-token verified continuations:

Mode	tok/s	Multiplier vs AR	Aggregate acceptance
Autoregressive	9.48	1.00x	—
MTP depth 1	16.72	1.76x	96.15%
MTP depth 2	29.67	3.13x	98.88%
MTP depth 3	24.64	2.60x	86.38%

Depth 2 is the verified recommendation. All depths generated the complete continuation and passed the runtime quality gate.

Vision was tested through the real MTPLX OpenAI-compatible multimodal endpoint. The model correctly returned both visible folder names in the supplied image.

Usage

bash

mtplx pull philipjohnbasile/Qwen3.6-27B-Refusal-Ablation-v2-MTPLX-6bit
mtplx serve \
  --model philipjohnbasile/Qwen3.6-27B-Refusal-Ablation-v2-MTPLX-6bit \
  --profile sustained \
  --depth 2

The checkpoint can also be opened by compatible mlx-lm tooling for autoregressive generation; the MTPLX sidecar and runtime contract provide the verified speculative path.

Build provenance

—Official Qwen/Qwen3.6-27B bf16 trunk.
—Locked candidate: global-d40-attn-late-300-mlp-broad.
—Direction index 40; layer-shaped attention and MLP intervention.
—MLX affine 6-bit body, group size 64.
—Projected bf16 MTP and vision sidecars.
—134 architecture-derived projection targets accounted for.
—20 immutable artifact files checked with zero hash mismatches.
—Final tensor payload: 23,649,238,436 bytes before Forge runtime metadata.

The repository includes the build receipt, locked candidate, quality receipt, holdout results, Heretic comparison outputs, and final summary.

Limitations and use

Refusal ablation can increase willingness to answer unsafe or inappropriate requests. It does not provide factuality, legality, or operational-safety guarantees. Deploy behind application-level policy, authorization, monitoring, and logging appropriate to your use case.

The behavioral datasets are now consumed evidence and must not be reused to select a future revision without a new independent holdout.

Credits

—Qwen for Qwen3.6-27B.
—p-e-w/Heretic for the KeywordRate reference protocol.
—MTPLX for the Apple-Silicon native-MTP runtime and Forge verification pipeline.

Apache-2.0, inherited from the official Qwen release.

Documentation provenance

ablation_build_receipt.json is an immutable build-time receipt. Its README entry hashes an 81-byte build placeholder, not this subsequently published model card. It should not be used as a checksum for the current README.