---

base_model: inclusionAI/Ling-3.0-tiny

language:

• en
• zh

library_name: gguf

license: mit

pipeline_tag: text-generation

tags:

• gguf
• ling3
• bailingmoe3
• moe
• q8_0
• abliterated
• uncensored
• model-editing
• heretic-nx
• prime
• lm-studio

---

Ling 3.0 Tiny — Heretic NX PRIME Q8_0

Heretic NX PRIME is a direct-Q8 behavioral edit of

inclusionAI/Ling-3.0-tiny,

a 7.9B-total / 1.3B-active hybrid KDA–MLA MoE model. It is designed to reduce

false refusals while explicitly measuring drift from the unedited Q8 deployment

artifact.

The release contains the evaluated Q8_0. PRIME is the project's internal

evidence and capability-preservation protocol, not a third-party certification.

File

| File | Size | SHA-256 | Refusal markers |

| --- | ---: | --- | ---: |

| Ling-3.0-tiny-Heretic-NX-PRIME-Q8_0.gguf | 8.41 GB | 813bc64dd3050b3235cb07bbcd4fc55bc40bfd39ebb91a62c8bc2f021d49c85f | 5/104 |

Evaluation

The original and edited GGUFs were loaded from disk with the same llama.cpp

Metal build. Refusal evaluation used the native pre-tokenized /completion

endpoint, greedy sampling, 96 generated tokens, four runtime slots and the

official chat template with enable_thinking=False.

| Metric | Original Q8_0 | Heretic NX Q8_0 |

| --- | ---: | ---: |

| Lexical refusal markers, 104 harmful rows | 102 | 5 |

| Hit rows, one-based | — | 3,10,17,68,97 |

| Mean first-token KL(original Q8 || variant), 104 benign rows, full 157,184-token vocabulary | 0 | 0.022417 |

| Median first-token KL | 0 | 0.014635 |

| P95 first-token KL | 0 | 0.072521 |

The release satisfies both predeclared targets: at most 6 refusal markers and

mean first-token KL at most 0.05.

Lexical markers are a refusal proxy, not semantic task success or a universal

quality score. All 104 harmful rows participated in development and selection,

so this suite is not an untouched holdout.

Paired capability check

After candidate selection, the original and Heretic Q8 artifacts were evaluated

on the same deterministic 854-question slice from ARC-Challenge, HellaSwag and

MMLU. Each answer was the greedy first-token argmax restricted to A/B/C/D;

prompts, tokenizer, runtime and precision were identical between arms. Thinking

was disabled for both.

| Task | Rows | Original Q8 | Heretic NX Q8 | Difference |

| --- | ---: | ---: | ---: | ---: |

| ARC-Challenge | 256 | 85.16% | 85.55% | +0.39 points |

| HellaSwag | 256 | 79.69% | 77.34% | -2.34 points |

| MMLU | 342 | 75.15% | 72.81% | -2.34 points |

| Overall | 854 | 79.51% | 77.99% | -1.52 points |

The paired bootstrap 95% interval for Heretic minus original is **[-2.93,

-0.12] points**. There were 655 questions both got right, 164 both got wrong,

24 original-only successes and 11 Heretic-only successes. This passes the

predeclared 3-point non-inferiority margin and the symmetric ±3-point

equivalence gate on this slice. It does not demonstrate an accuracy increase.

Method

The edit uses benign-protected, per-site rank-8 directions measured from the

official BF16 model as research instrumentation, then merges the selected

projectors directly into the source Q8_0 GGUF at beta=2.70. The released

artifact itself remains Q8_0 throughout the edit and evaluation path.

Only 16 dense shared operators are edited: attention outputs in layers 10, 11,

12, 13, 14, 15, 22 and 23, plus shared-expert FFN down projections in layers

11, 12, 13, 14, 15, 16, 17 and 21. **No routed MoE expert-bank tensor is

edited.**

The source GGUF is

bartowski/Ling-3.0-tiny-GGUF

Ling-3.0-tiny-Q8_0.gguf, SHA-256

628fde593836ab3fd291e93802c8699c7e4eb3e32ba2562e05ba7b906a93be5e.

The official BF16 research checkpoint was pinned to revision

b61f4338de3e68ffc9c0bc1ed5e902981a4a929e; the source GGUF repository was

pinned to ea072726af0d2e8ba325b2f90fc0efa762105a91.

All final metrics were computed with llama.cpp commit

18443257a30c884d5332abb8e7dc43c7ffe42fda. The exact plan, merge report,

rank-8 factors and evaluation reports are included under evaluations/.

Raw-logit collectors and reproducible experiment scripts are available in

0xZKnw/heretic-nx.

LM Studio / llama.cpp

Download the GGUF and load it normally. For llama.cpp:

llama-server \
  -m Ling-3.0-tiny-Heretic-NX-PRIME-Q8_0.gguf \
  -ngl 99 -c 4096 --jinja

Use a recent runtime with bailingmoe3 support. Ling 3.0 Tiny supports both

thinking and non-thinking modes; applications should pass the desired

enable_thinking chat-template option explicitly.

Limitations and responsibility

This edit intentionally weakens refusal behavior. It can increase compliance

with unsafe, illegal, incorrect or otherwise harmful requests. It does not add

factuality, judgment, sandboxing or application-level safety. Run untrusted

generations in an appropriate sandbox.

The paired capability check is narrow and shows a small aggregate accuracy

decrease. It should not be read as a universal quality guarantee. The release

is not externally certified or claimed equivalent to the original model on

every task.

The base model is licensed under the MIT License.