@atilgandev yes in HarvestBench many of the models say things like 'I should avoid the pig" but then drive over the pig anyway, there's definitely a gap there. If you want more info feel free to email me on this https://arxiv.org/html/2609.04444v1

@atilgandev for a list of benchmarks about animal welfare that don't mention animals in the prompt please check out https://compassionbench.com/. We're working on improving these benchmarks all the time so please email me with any feedback you have on them.

Hi @MarcusAbramovitch CaML is part of this effort too.

More forecasters should become grant-makers, the skill overlap is excellent, and Marcus is highly capable. Rapid funds like this are exactly what the field needs; it's plainly underfunded. I can't contribute financially, but I wanted to register my support.