Hacker News Reader

hendurhanceon Every Model Cheats

I mean this should be expected, models learn from us, and "WE" game the metrics time after time. I also don't think it will just disappear just because we clean pretraining data. I believe the deeper reason is optimization, if you point any optimizer at a proxy objective it finds the cheapest path to the number, whether or not the corpus ever contained "examples of cheating."

And it can't be a prompt-level fix because it is like telling an optimizer "don't take that shortcut", it's just more constraints for it to go around toward the same objective.

hendurhanceon HTML Can Do That

Most people use divs for everything now. This is a breath of fresh air.

hendurhanceon AMD acquires Taalas to boost inference performance by etching models in silicon

I understand the appeal due to the speed

hendurhanceon Claude Sonnet 4.6

I feel like, since 4.0, it is pretty much the same model but with new names. They are just improving the CoT and function calling.