Post

Andrew White 🐦‍⬛
@andrew.diffuse.one

HLE has recently become the benchmark to beat for frontier agents. We at FutureHouse took a closer look at the chem and bio questions and found about 30% of them are likely invalid based on our analysis and third-party PhD evaluations. 1/7

Andrew White 🐦‍⬛
@andrew.diffuse.one

The design process of HLE required the questions to be unanswerable by contemporary LLMs. That lead to many gotcha style questions like the one below. It’s a trick question – in 2002, a few atoms of a group 18 element Oganesson were made for a few milliseconds. 2/7

Andrew White 🐦‍⬛
@andrew.diffuse.one

It’s a clever question. But it’s not really about frontier science. Multiple papers have shown that Oganesson is not a gas (it’s predicted to be semiconducting solid), it’s not noble (it’s reactive), and it isn’t included in any "terrestrial matter" tables of noble gases. 3/7

0 reposts0 likes

Andrew White 🐦‍⬛
@andrew.diffuse.one

The HLE rubric wanted questions to have “objectively correct, univocal” ground-truth answers. You can find multiple peer-reviewed papers that contradict the statement "Oganesson was the rarest noble gas in 2002 as a percentage of terrestrial matter" 4/7

Andrew White 🐦‍⬛
@andrew.diffuse.one

We reviewed 150 of the questions in the chem and bio and found about 30% have peer-reviewed papers contradicting their ground-truth answers. Issues include confusion of species with orders, misreading of FDA guidelines, etc. All our notes are public. 5/7

Andrew White 🐦‍⬛
@andrew.diffuse.one

We make evals at FutureHouse. It’s hard and it sucks. It’s also now the bottleneck, as we scratch the boundary of human ability. HLE was a huge effort and made many good questions and we hope this analysis stimulates review of the other HLE categories and improvements 7/7