Hackernews posts about ARC-AGI-1

1. Samsung released a 7M model that achieved 45% on ARC-AGI-1 (arxiv.org)
   34 points by chintler 12 months ago | 12 comments
2. F. Chollet: We were able to reproduce the findings of the HRM paper on ARC-AGI-1 (twitter.com)
   5 points by Anon84 about 1 year ago | discuss
3. Inkling is now the highest-scoring open-weight model on both ARC-AGI 1 and 2 (twitter.com)
   3 points by thebricklayr 2 months ago | 1 comments
4. 150M-parameter reasoning model sets new cost-accuracy frontier on ARC-AGI-1 (huggingface.co)
   3 points by meander_water about 1 month ago | discuss
5. 44% on ARC-AGI-1 in 67 cents (github.com)
   1 points by evilmathkid 7 months ago | 1 comments
6. ARC-AGI-1 and 2 LEADERBOARD (arcprize.org)
   1 points by doener 7 months ago | 1 comments
7. TranscendPlexity: 540/540 ARC-AGI-1/2/3, 13 tasks with 0% AI solve rate, solved (github.com)
   1 points by wormsWorld 4 months ago | discuss
8. Poetiq announces new SOTA on the ARC-AGI-1 and 2 (poetiq.ai)
   1 points by fofoz 10 months ago | discuss