INDEX

Explanations

the start and end of structured or formatted data sections, particularly in markup or programming contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 12-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_12/width_65k/average_l0_72

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.12.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 suicidio

-0.44

 diamètre

-0.43

={<

-0.40

周泽楷

-0.40

rav

-0.40

-0.39

 tuer

-0.38

метил

-0.38

éről

-0.38

 radical

-0.38

POSITIVE LOGITS

 CreateTagHelper

1.22

 BoxDecoration

1.11

 ***!

1.09

enumi

1.05

 imageNamed

1.05

 חיצוניים

1.00

RectangleBorder

0.99

\{\\

0.99

 مشين

0.98

ValueStyle

0.97

Activations Density 0.002%

the start and end of structured or formatted data sections, particularly in markup or programming contexts

No Comments

No Known Activations