Re:Launch HN: General Instinct (YC P26) – Frontier models on...
by BoorishBears on Friday June 05, 2026 @02:14PM
(#48416203)
I like the technique described here around distillation to recover from quantization, but I don't understand why we keep performing lossy compression on LLMs then using benchmarks that were nearly saturated before post-training to measure the effects.
You could erase the gains from literally half the compute going into some of these recent models and barely make a dent in MMLU-Pro and GPQA-D.
You could erase the gains from literally half the compute going into some of these recent models and barely make a dent in MMLU-Pro and GPQA-D.