RVN Qwen3.8-Flash-Next Abliterated — native Halogen (.hgn)

Unofficial community build: a native Halogen checkpoint repacked from
0bserverx/RVN-Qwen3.8-Flash-Next-Abliterated-Uncensored-GGUF (IQ4_XS, 8 shards)
with halogen-flash-server:0.15.3. Serves directly on halogen-flash-server >= 0.15.3 —
no GGUF, no sidecar files needed.

File

• rvn-q38fn-abliterated.hgn — 101,995,332,160 bytes
• SHA256: c46b0da5f8c661c5a086e03711416aa3cd866467fc4d441bc5639e8b46b5c130
• HGN1 container, 1198 tensors. FP8 PLE ngram table and MTP draft head are folded in.

Provenance

• Base model: Qwen/Qwen3.8-Flash-Next, used under its original license terms;
  this repository follows those terms.
• Abliteration weights: 0bserverx/RVN-...-GGUF (IQ4_XS) — created independently by
  the community author credited in the source repo. No retraining here.
• This repo contains only repair + container conversion of the source weights
  (no requantization; weights pass through unchanged).

Indexer repair

The conversion tooling that produced this GGUF exhibits systematic defects in the
sparse-attention (QSA) indexer: misaligned index_qk_proj values, zeroed or shifted
layernorms, and in some builds F16 stored where BF16 is expected. These defects are
silent (the model loads and converses normally) but degrade sparse-layer retrieval.

All indexer values below were repaired from the pristine base model
(Qwen/Qwen3.8-Flash-Next):

1. All 12 index_qk_proj tensors set to base-model BF16 (bit-exact).
2. F16 -> BF16 dtype correction on the indexer tensors.
3. Indexer layernorms validated against the canonical GGUF convention (stored gamma = 1 + w).
4. Container conversion with flash_serve --repack (0.15.3), folding in the official
   MTP head and the engine's FP8 PLE ngram table.

Verification receipts

• halogen-tools verify (0.15.3): PASS — 1198 tensors, 94.99 GiB, v2,
  model_id qwen3.8-flash-next-gguf; header, table, layout, payload sizes, checksums,
  codebooks and scales OK; geometry checked on 1197 tensors (1 with no rule).
• Teacher-forced perplexity, 6,105-token mixed corpus (prose/code/dialog), chunk 1024,
  identical tokenizer and corpus for every run:
  • this file: PPL 3.511 (mean NLL 1.2558)
  • official qwen38-flash-next-v2.hgn anchor: PPL 3.307 (mean NLL 1.1960)
    Uniform small deltas across all length bands are consistent with quantization-level
    difference rather than structural damage.
• Live serve on 0.15.3: chat completions OK, instruction-following exact,
  ~33 tok/s decode with MTP drafting, ~107 tok/s prefill (Strix Halo, gfx1151).
• SHA256 above re-verified against the file that was served.

Usage

Point HALOGEN_CHECKPOINT at the file with halogen-flash-server >= 0.15.3
(tokenizer mounted at /tokenizer). The ngram table and draft head are inside the file.

Access & terms

This repository is gated: please state your intended use when requesting access.

• Unofficial research build. Not affiliated with the Qwen team.
• Behaviour (including "abliterated/uncensored" outputs) is inherited unchanged from
  the source GGUF and its license. Evaluate at your own discretion; no warranty.

Notes

• Repack is deterministic: an independent second repack of the same inputs is byte-identical.