“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.png -- Do not miss in the article the "Related work" chapter, which also offers other interesting budding directions.”
BDH-CQ: In-Context Learning with Recurrent Latent Reasoning
We introduce BDH-CQ, a reasoning model that combines in-context learning with recurrent latent reasoning. Inputs presented at inference time continuously update the model's recurrent memory; the model then solves a query through iterative computation in a high-dimensional latent space, without verbalizing its intermediate reasoning. We evaluate the model on the public ARC-AGI-1 evaluation set and use controlled ARC-like interventions to study what it learns from demonstrations, how consistently it applies an inferred transformation, and which concepts remain difficult. A 150M-parameter configuration reaches 29.5% pass@2 at a computed inference cost of \$0.0007 per task. This operating point breaks through the previously reported ARC-AGI-1 cost-accuracy Pareto frontier, establishing a new state of the art in benchmark cost efficiency.
候选层:至少 2 个不同的人、在 2 个不同月份提到过,还没到主书架的 3 人 3 个月。没有做立场标注,所以只写“被提及”。
HN 书评
推荐在前,同类按时间 · 每篇最多保留 20 条,统计用全部提及- 150M-parameter reasoning model sets new cost-accuracy frontier on ARC-AGI-1 ↗meander_water · 2026-08-12 · 1 分
- New model BDH-CQ costs $0.007 per task 11x less than OpenAI Luna even w 80% off ↗wordpad · 2026-08-13 · 4 分
- BDH-CQ: In-Context Learning with Recurrent Latent Reasoning ↗flaburgan · 2026-08-22 · 2 分
- 30% on AR-AGI-1 at $0.0007 per task ↗randomizedalgs · 2026-08-29 · 3 分
- BDH-CQ: In-Context Learning with Recurrent Latent Reasoning ↗mdp2021 · 2026-09-02 · 3 分