moravianai/llm-weight-compression-evidence

llm-weight-compression: evidence Per-window evaluation data and raw diagnostics behind every number in rosecky/llm-weight-compression-public: a pre-registered study of refining GPTQ's integer codes against the output of the enclosing attention/MLP module instead of the single layer, on Qwen2.5-0.5B, Qwen3-0.6B, Llama-3.2-1B and Qwen2.5-1.5B at exactly 3.25 bits per weight, plus rate–distortion, estimation-budget and 2-bit vector-quantized points on Qwen2.5-0.5B. The long-form… See the full description on the dataset page: https://huggingface.co/datasets/moravianai/llm-weight-compression-eviden

<iframe src="https://silytics.si/embed/item/c95a7fd1-6c9a-4ad5-bba9-34d0e18c88f0" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 11 November 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.