Hacker News new | ask | show | jobs
by WarmWash 53 days ago
The open models only give the SOTA models a run for their money on gameable benchmarks. On the semi-private ARC-AGI 2 sets they do absolutely awfully (<10% while SOTA is at ~80%)

It might be too expensive, but I would be interested in the benchmarks for the current crop of SOTA models.
1 comments
roenxi 53 days ago
Have the open models been tried? When I look at the leaderboard [0] the only qwen model I see is 235B-A22B. I wouldn't expect an MoE model to do particularly well, from what I've seen (thinking mainly of a leaderboard trying to measure EQ [1]) MoE models are at a distinct disadvantage to regular models when it comes to complex tasks that aren't software benchmark targets.

[0] https://arcprize.org/leaderboard

[1] https://eqbench.com/index.html

link
WarmWash 52 days ago
There is GLM 5 and kimi 2.5 (which gets 11.8%, but I digress)
link