INDEX

Explanations

jupyter

np_max-act · gemini-2.0-flash

New Auto-Interp

Configuration

andyrdt/saes-qwen2.5-7b-instruct/resid_post_layer_15/trainer_1

Dataset (Dashboard)

Various

Features

131,072

Data Type

float32

Hook Name

blocks.15.hook_resid_post

Architecture

standard

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ær

-0.07

дар

-0.07

 Compatibility

-0.07

ASHINGTON

-0.07

-0.06

 USSR

-0.06

🎐

-0.06

 stealth

-0.06

 kunne

-0.06

.Width

-0.06

POSITIVE LOGITS

 Represent

0.08

\Input

0.07

activation

0.07

 Jenkins

0.07

 numberOf

0.07

 heatmap

0.07

 predomin

0.07

 Staples

0.07

 Baton

0.07

 ва

0.07

Activations Density 0.008%

jupyter

No Comments

No Known Activations