BDH-CQ: In-Context Learning with Recurrent Latent Reasoning · HackerPapers discussions

BDH-CQ: In-Context Learning with Recurrent Latent Reasoning — mentioned by mdp2021

Wed, 02 Sep 2026 07:07:07 GMT

A Reasoning Model that as > a 150M-parameter BDH-CQ reaches 29.5% pass@2 on the public ARC-AGI-1 evaluation set at a computed $0.00070 per task Comparison: https://arxiv.org/html/2608.09888v1/figures/leaderboard_295.... -- Do not miss in the article the "Related work" chapter, which also offers other interesting budding directions.

Paper and discussion history. A mention is not an endorsement.

BDH-CQ: In-Context Learning with Recurrent Latent Reasoning — HN submission

Wed, 02 Sep 2026 07:02:54 GMT

BDH-CQ: In-Context Learning with Recurrent Latent Reasoning

3 points; 1 discussion reply on HN. Paper and source history. A submission is not an endorsement.

BDH-CQ: In-Context Learning with Recurrent Latent Reasoning — HN submission

Sat, 29 Aug 2026 13:35:49 GMT

30% on AR-AGI-1 at $0.0007 per task

3 points; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.

BDH-CQ: In-Context Learning with Recurrent Latent Reasoning — HN submission

Sat, 22 Aug 2026 08:59:01 GMT

BDH-CQ: In-Context Learning with Recurrent Latent Reasoning

2 points; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.

BDH-CQ: In-Context Learning with Recurrent Latent Reasoning — HN submission

Thu, 13 Aug 2026 17:48:38 GMT

New model BDH-CQ costs $0.007 per task 11x less than OpenAI Luna even w 80% off

13 points; 1 discussion reply on HN. Paper and source history. A submission is not an endorsement.

BDH-CQ: In-Context Learning with Recurrent Latent Reasoning — HN submission

Wed, 12 Aug 2026 10:21:52 GMT

150M-parameter reasoning model sets new cost-accuracy frontier on ARC-AGI-1

3 points; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.