Heretic v2: uncensored Qwen 3.8 125B quant tuned for reverse engineering with activation-energy bit allocation
jjusko20 · reddit · 2026-10-30
A Reddit user released Qwen 3.8 Swift 1.5 Flash Heretic v2 (uncensored, 125B MoE GGUF, 96GB VRAM), built with jeffpeng3's v2 heretic weights. The quant uses a custom 400k-token importance matrix weighted toward chain-of-thought, system prompts, agentic tool use, and an 8.8% reverse-engineering domain (Ghidra, deobfuscation, anti-debug, Frida/eBPF tracing), fused 70/30 with AtomicChat's general-purpose matrix. Bit allocation follows activation-energy auditing: layers 36–47 down-projections at near-lossless 8-bit, layer 0 attention gate kept uncompressed, layer 2's n-gram merge point protected. Recipes are public.
Related event: Uncensored, reverse-engineering-tuned Qwen 3.8 quant released(2 posts)→
More from Models
- Gemini 4 Carbon takes #1 on Artificial Analysis, Anthropic S-1 in 60+ charts — altryne · 2026-10-30
- Leaker shows OpenAI shipped his ChatGPT texting feature ideas two days later — flowersslop · 2026-10-30
- First true omni model runs fully in-browser: 280 tok/s with WebGPU — victormustar · 2026-10-30
- Hidden ChatGPT checkout text reveals shared usage limits across ChatGPT and Codex starting December — btibor91 · 2026-10-30
- Grok 4.7 ranks least sycophantic model in AA-PersonalityBench, Slop score falls from 68 to 6 — XFreeze · 2026-10-30
- Epoch AI: GLM 5.3 autonomously exploits unpatched flaws to compromise six Firefox versions — Miles_Brundage · 2026-10-30