←News detail

English text is being prepared.Updated Nov 1 11:24

SenseTime Open-Sources SenseNova Loop-9B-Preview, Claiming It Rivals 27B-Parameter Models

English text is being prepared.

AI summary

SenseTime announced on October 30 the open-sourcing of SenseNova Loop-9B-Preview. The model adopts a Looped Transformer architecture and expands effective computational depth through recurrent computation; the company claims that at only 9B parameters it can rival 27B-parameter models. The model retains the original decoder modules of 4 layers each at the beginning and end, converts the middle layers into parameter-sharing recurrent computation modules, and introduces mHC (Manifold-Constrained Hyper-Connections) to enhance feature interaction. The company claims that it surpasses Qwen3.5-27B on evaluations such as WorkBuddy Bench, RaccoonPPT Bench, and AutomationBench, and that in ablation experiments GPQA Diamond and TeamBench improved by 4.42 points and 13.34 points, respectively.

Why read this

AI score 60

SenseTime uses recurrent computation to increase the effective computational depth of a 9B model and provides comparative evaluations against 27B models, offering a basis for judging the small-parameter-model approach.

Topics

Tags

Body

商汤开源 SenseNova Loop-9B-Preview,宣称媲美 27B 参数大小模型

2026-11-01 09:02 IT之家 - 沁沧(实习)

IT之家 11 月 1 日消息,商汤科技 10 月 30 日宣布开源商汤日日新 SenseNova Loop-9B-Preview。

官方表示,该模型采用 Looped Transformer 架构,引入循环计算、扩展模型的有效计算深度,并探索更充分的信息交互机制,提升模型在复杂任务上的处理能力,在仅 9B 的有限参数规模下,媲美 27B 参数大小模型。

SenseNova Loop-9B-Preview 以 9B 尺度模型为基座,保留首尾各 4 层原始解码器模块,将中间层改造为参数共享的循环计算模块,通过多轮迭代扩展有效计算深度。在此基础上,模型进一步探索循环计算中的信息传递机制,尝试引入 mHC(Manifold-Constrained Hyper-Connections),利用多路残差连接与受约束的信息混合,增强特征交互并维持信号传播的稳定性。

据介绍,在当前评测设置下,模型在代码生成、办公场景和智能体任务等多个方向上均优于其他模型,并在 WorkBuddy Bench、RaccoonPPT Bench 和 AutomationBench 等评测中进一步超过 Qwen3.5-27B。这些结果初步展现了通过扩展有效计算深度,提升小参数模型复杂任务处理能力的潜力。

商汤进一步通过消融实验评估了 mHC 的实际收益。采用该机制后,模型取得明显的性能提升,其中 GPQA Diamond 和 TeamBench 分别提高了 4.42 分和 13.34 分,展现出在复杂推理任务以及多步骤 Agent 任务中的应用潜力。后续还将继续探索 DSA、自适应迭代次数等机制,以提升模型的推理效率与任务表现。

IT之家附开源链接:

广告声明:文内含有的对外跳转链接(包括不限于超链接、二维码、口令等形式),用于传递更多信息,节省甄选时间,结果仅供参考,IT之家包含外链的文章均包含本声明。

Source

English text is being prepared.