AI 模型和平台
Cognition 发布 SWE-2 Flash for更快，下层编码 Agents
在 Google 上的首选来源中添加 Unite.AI
2026年11月4日Cognition 宣布 SWE-2 Flash报道，编码模型运行速度是其更大的SWE-2兄弟姐妹的三倍，价格是五分之一。 它可以在Devin桌面和CLI中找到，目的是常规的软件-agent工作，而不是统一的基准测试领导。
计划的工作包括自动化，代码审查，特征开发，以及文档更新，免费使用至2026年11月30日。 Flash拥有2,760亿总参数和120亿活性参数，匹配了思维机器实验室的Inkling-Small，混合型专家基础模型Cognition. 2026年7月30日"思机发布"墨林-小的全重。
基准测试收益和贸易业务
Cognition报告以下评价分数，显示对SWE-2的Inkling-Small取得了巨大收益，但结果好坏参半：
在Cognition的比较中，Flash边缘SWE-2在FrontierCode2.0上，而SWE-2则在终端-Bench 4.0和DeepSWE 1.1.上得分较高。 结果将Flash作为成本和速度的替代物，而不是其更大的兄弟会的全面替代物。
Cognition 齐克0Z 在它报告的主要评价上，闪亮的成绩超过了所有美国开放量级模型，包括三倍大的模型。 它的生成速度插图对方法很具体：每个模型自己的终端-Bench 4.0gsea-蛋白质组运行以模拟速率重播，Flash每秒200令令牌，SWE-260令牌，笔录缩短到同等长度。
中级培训和强化学习
Cognition 称这是其食谱首次在强化学习前设置了大规模的中训练阶段。 为了选择数据混合物，团队整理了一个被搁置的编码数据集，意在代表真正的软件工程情景。 将候选混合物与达到基准运行的验证损失所需的每个标志进行比较，使标志性效率成为一项选择标准。
最终的混合体中包含从科学文章、博客、代码和其他来源抽取的几兆个符号。 培训分三个阶段进行，因此后来出现了最稀少、质量最高的数据。 Cognition报告说，这些阶段的搁置验证损失以稳定的对数线率有所改善。
随后进行了大规模强化学习。 Cognition将SWE-2食谱的改进大部分归结为质量更高、种类更多样化的任务，而不是食谱调整。 该小组审计了任务说明、环境设置和核查人员，并增加了任务类型，目的是为真正的Devin工作流程建立更广泛的推理和解决问题。
限制的重量访问
Flash的体重尚未公开分享。 Cognition正在为信任的行业伙伴和研究实验室建立一个安全共享方案，并许诺提供更多细节。 计划访问将使已核准的合作伙伴微调能够使用该模型，并在敏感的客户控制环境中运行，包括精密基础设施，同时研究人员可以研究行为和可能的滥用。
Cognition报道称，针对网络安全训练的融资 Expectit Bench从Inkling-Small的11.3%到34.8%，但警告同样的能力可以帮助恶意攻击者，促使决定不公开权重。
Cognition的2026年10月26日安全帖子描述AI加速了攻击者和维权者的工作，并讨论了2026年10月21日披露的PHP和Node.js远程码执行漏洞。
该版本将即时用途与更广泛的接入区分开来。 在Devin内部，Flash是报道的日常工程工作的成本和速度选择。 除此产品外，其对企业和研究人员的价值取决于计划中的程序能否在Cognition设定的安全限制下交付微调和客户控制的部署。
本文由 AI图鉴自动抓取并翻译，版权归原作者所有；如需查看原文请点下方链接。
AI Models & Platforms
Cognition Launches SWE-2 Flash for Faster, Lower-Cost Coding Agents
Add Unite.AI to your preferred sources on Google
Cognition announced SWE-2 Flash on November 4, 2026, and reports that the coding model runs at three times the speed and one-fifth the price of its larger SWE-2 sibling. Available in Devin Desktop and CLI, it is aimed at routine software-agent work rather than uniform benchmark leadership.
The intended jobs include automations, code review, feature development, and documentation updates, with free use offered through November 30, 2026. Flash has 276 billion total parameters and 12 billion active parameters, matching Thinking Machines Lab’s Inkling-Small, the mixture-of-experts base model Cognition adapted. Thinking Machines released Inkling-Small’s full weights on July 30, 2026.
Benchmark Gains and Trade-Offs
Cognition reports the following evaluation scores, which show substantial gains over Inkling-Small but mixed results against SWE-2:
In Cognition’s comparison, Flash edges SWE-2 on FrontierCode 2.0, while SWE-2 scores higher on Terminal-Bench 4.0 and DeepSWE 1.1. The results frame Flash as a cost-and-speed alternative rather than an across-the-board replacement for its larger sibling.
Cognition says Flash outperforms all American open-weight models on the leading evaluations it reports, including models three times larger. Its generation-speed illustration is specific about the method: each model’s own Terminal-Bench 4.0 gsea-proteomics run is replayed at simulated rates of 200 tokens per second for Flash and 60 for SWE-2, with transcripts shortened to equal length.
Mid-Training and Reinforcement Learning
Cognition says this is the first time its recipe has placed a large-scale mid-training stage before reinforcement learning. To choose the data mixture, the team curated a held-out coding dataset intended to represent real software-engineering scenarios. Candidate mixtures were compared by how many tokens each required to reach the baseline run’s validation loss, making token efficiency a selection criterion.
The final mixture contained a few trillion tokens drawn from scientific articles, blogs, code, and other sources. Training proceeded in three stages so the scarcest, highest-quality data appeared later. Cognition reports that held-out validation loss improved at a stable log-linear rate across those stages.
Large-scale reinforcement learning followed. Cognition attributes most of the improvement over its SWE-2 recipe to higher-quality and more diverse tasks rather than to recipe adjustments. The team audited task instructions, environment setup, and verifiers, and added task types intended to build broader reasoning and problem-solving for real Devin workflows.
Restricted Weight Access
Flash’s weights are not being shared publicly yet. Cognition is building a secure sharing program for trusted industry partners and research labs, with further details promised. The planned access would let approved partners fine-tune the model and run it in sensitive customer-controlled environments, including on-premise infrastructure, while researchers could study behavior and potential misuse.
Cognition reports that targeted cybersecurity training raised ExploitBench from Inkling-Small’s 11.3% to 34.8%, but warns that the same capabilities could assist malicious attackers, prompting the decision to withhold public weights.
Cognition’s October 26, 2026 security post describes AI as speeding up work for both attackers and defenders and discusses the PHP and Node.js remote-code-execution vulnerabilities disclosed on October 21, 2026.
The release separates immediate utility from broader access. Inside Devin, Flash is a reported cost-and-speed option for everyday engineering work. Beyond that product, its value to enterprises and researchers depends on whether the planned program can deliver fine-tuning and customer-controlled deployment under the security restrictions Cognition has set.
Auto-fetched and translated by AI Atlas. All rights belong to the original author.