Ablated Gemma 4 E2B variants slashed refusals to sub‑1% in extensive benchmark
A new GitHub repo details a comprehensive benchmark of 13 ablated Gemma 4 E2B variants, evaluating weight analysis, KL divergence, HarmBench safety, and eight downstream tasks after 44 GPU‑hours on an RTX 5090. The results show that norm‑preserving biprojection plus Expert‑Granular Abliteration cuts refusal rates to under 1 % while keeping low KL for larger models.