PulseAugur
EN
LIVE 08:23:48

New reasoning model BDH-CQ achieves state-of-the-art cost efficiency on ARC-AGI-1

By PulseAugur Editorial · [2 sources] ·

Researchers have developed BDH-CQ, a novel reasoning model that integrates in-context learning with recurrent latent reasoning. This model updates its memory with inference-time inputs and iteratively computes solutions in a latent space without verbalizing intermediate steps. Evaluated on the ARC-AGI-1 benchmark, a 150M-parameter version achieved 29.5% pass@2 at a low cost, establishing a new state-of-the-art for cost efficiency on this benchmark. AI

IMPACT Establishes a new state-of-the-art in cost efficiency for reasoning tasks, potentially enabling more accessible advanced AI capabilities.

RANK_REASON The cluster describes a new research paper detailing a novel model and its performance on a benchmark. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.NE (Neural & Evolutionary) →

AI-generated summary · Google Gemini · from 2 sources. How we write summaries →

New reasoning model BDH-CQ achieves state-of-the-art cost efficiency on ARC-AGI-1

COVERAGE [2]

  1. arXiv cs.NE (Neural & Evolutionary) TIER_1 English(EN) · Richard Zhong ·

    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 …

  2. arXiv stat.ML TIER_1 English(EN) · Bj\"orn Engdahl, Adrian Kosowski, Jan Chorowski, Zuzanna Stamirowska, Przemys{\l}aw Uzna\'nski, Junlin Jiang, Rohan Phadke, Remigiusz Kinas, Richard Zhong ·

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

    arXiv:2608.09888v1 Announce Type: cross Abstract: 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 i…

RELATED ENTITIES

RELATED TOPICS