News for hackers. Stuff that matters.
Re:Launch HN: General Instinct (YC P26) – Frontier models on... by rohansood15 on Friday June 05, 2026 @01:40PM (#48415809)
Have you benchmarked against other 3-bit dynamic quants like Unsloth? I am sorry but this framing against a full precision, newer, smaller MoE just seems misleading. Also, Gemma-4-26B-A4B is not the SOTA for edge. Even at launch, that would be the 31B.
Re:Launch HN: General Instinct (YC P26) – Frontier models on... by guanming0717 on Friday June 05, 2026 @01:44PM (#48415857)
Yes I did, with other SOTA quant methods like HQQ, AWQ etc. You can find more info in our blog :) https://general-instinct.com/blog/frontier-moe-sub-4-bit
Re:Launch HN: General Instinct (YC P26) – Frontier models on... by rohansood15 on Friday June 05, 2026 @01:58PM (#48416016)
I can't find it. Can you state your performance versus comparable 3-bit quantization from Unsloth/Bartowski? Edit: I appreciate that you seem to have open-sourced the quantization pipeline. This is not to question your work, but to understand where the outputs stand relative to the SoTA for quantization.