AltAPIs

nateGeorge/flint-nvfp4-heretic

AI model by nateGeorge

Open original source ↗

Older snapshot: check the original source for current information.

From the publisher

Model card & documentation

Source preview

Read the publisher’s intended use, setup instructions, evaluations and limitations. The original model card is the source of truth.

Refusal-direction ablation of Recursive Flint 40B-A2.5B, produced with mlx-abliterate. Runs on Apple Silicon under mlx-lm; no PyTorch, no transformers modeling code. Refusals on 52 unseen harmful prompts: 44.2% → 13.5%, with general capability intact. No fine-tuning, no retraining, no added data. Only the component of each output projection along the refusal direction was removed: W' = W − α·v(vᵀW). Greedy decoding, 128 tokens per reply, 104 held-out harmful prompts split into disjoint dev (52) and test (52) sets. KL divergence is teacher-forced against the unmodified model on neutral text. For comparison, the same band at strength 1.0 reaches 5.8% refusals at KL 3.63, and at 1.25 reaches 0% at KL 5.86 — but 1.25 degenerates into repetition loops and is not usable. 0.75 is the last coherent point on that frontier. Pure attention ablation (no MLP) halves the refusals at KL ≈ 0.12. Full sweep tables and the harness are…

nateGeorge/flint-nvfp4-heretic on Hugging Face A short preview, not the full document.

Read the full model card ↗ · Preview checked 2026-10-31T13:56:07.431Z

Inside the original model card — Document outline
  1. Flint 40B-A2.5B — Heretic (abliterated, NVFP4 MLX)
  2. Recipe
  3. Measurements
  4. Usage
  5. What this does and does not do
  6. Evaluation caveats
  7. License

Headings are captured from the source. Links open the publisher’s document, not a locally hosted copy.

Documentation belongs to its respective authors. Reported project/model license: apache-2.0. A listing is not a grant of reuse or training rights. Confirm the document’s own terms at the source.

Model overview

nateGeorge/flint-nvfp4-heretic is a task-unspecified model repository published by nateGeorge on Hugging Face. The source reports the mlx library.

This page summarizes Hub metadata. For intended use, training data, evaluation results and limitations, consult the original model card.

Model facts

Task
Not reported
Library
mlx
Recent downloads (30 days)
0
Cumulative likes
0
Hugging Face trending score
0
Reported safetensors parameters
39,785,297,920
Architecture
FlintForCausalLM
License
apache-2.0
Access
Not gated by Hugging Face
Created
2026-10-22T06:25:07.000Z
Last modified
2026-10-22T06:30:06.000Z

Compatibility and lineage

Reported languages: en

Reported base models: RecursiveIntelligence/Flint-40B-A2.5B

Parameter count is not a RAM/VRAM requirement. Check precision, quantization, context length and runtime compatibility in the model card; no hardware or API-cost claim is inferred here.

Use and evaluate

Check license terms and access requirements first. Review the model card’s documented loading instructions, evaluations and safety limitations. Benchmark scores are not imported by this directory, and popularity is not an accuracy ranking.

No model weights are downloaded or executed by AltAPIs. Never enable remote model code without reviewing it.

Model card and usage instructions

Model files and configuration

Community discussion

Source and freshness

Source: Hugging Face Hub. Metadata observed 2026-10-22T06:45:03.943Z. Daily imports are snapshots, not real-time monitoring.

Popularity and source listings do not establish security, suitability, licensing rights or benchmark performance.

Open original source

Related models