返回
原文
Deedy@deedydas
51AI 编辑部评分，满分 100
2026-08-05 00:34· 27分钟前
跳到正文
AI 摘要

Goodfire 的 AI 研究平台 Silico 正式公开可用，支持在 frontier 规模下解释和训练模型。Deedy Das 实测称，Silico 能对视觉 Transformer（ViT）逐头可视化 ImageNet 图像的 196x196 注意力矩阵，并生成动画弧线展示 patch 间关系，是目前唯一能一次性完成此类综合 AI 研究的模型或产品。

We finally have a product that helps with frontier AI research.

Frontier LLMs are a） nerfed at AI research b） lack immediate GPU availability and c） lack access to the right skills and research material.

"Teach me how vision models see"

Goodfire Silico takes a vision transformer （ViT）， loads ImageNet images， and for each head， visualizes the 196x196 matrix by creating animated arcs across patches.

Haven't found another model or product that can do this level of comprehensive AI research， much less oneshot it.

GoodfireSilico, the platform for ambitious AI research, is publicly available today. AI is advancing fast. The tools to understand it need to advance even faster. Silic...
产品更新数据/训练
在 X 查看原推导出 Markdown
返回
51导出 Markdown
Deedy · @deedydas · X·2026-08-05 00:34·27分钟前
在 X 看原推· x.com
AI 摘要

Goodfire 的 AI 研究平台 Silico 正式公开可用，支持在 frontier 规模下解释和训练模型。Deedy Das 实测称，Silico 能对视觉 Transformer（ViT）逐头可视化 ImageNet 图像的 196x196 注意力矩阵，并生成动画弧线展示 patch 间关系，是目前唯一能一次性完成此类综合 AI 研究的模型或产品。

We finally have a product that helps with frontier AI research.

Frontier LLMs are a） nerfed at AI research b） lack immediate GPU availability and c） lack access to the right skills and research material.

"Teach me how vision models see"

Goodfire Silico takes a vision transformer （ViT）， loads ImageNet images， and for each head， visualizes the 196x196 matrix by creating animated arcs across patches.

Haven't found another model or product that can do this level of comprehensive AI research， much less oneshot it.

GoodfireSilico, the platform for ambitious AI research, is publicly available today. AI is advancing fast. The tools to understand it need to advance even faster. Silic...
产品更新数据/训练
在 X 查看原推x.com