INDEX

Explanations

instances of directional or positional movements and actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 2-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_2/width_16k/average_l0_67

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.2.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 '\\;'

-0.71

webElement

-0.70

 onPostExecute

-0.65

出版年

-0.64

rrggbb

-0.63

 ModelExpression

-0.63

دانشنامهٔ

-0.62



-0.59

 consultato

-0.59

aarrggbb

-0.59

POSITIVE LOGITS

 Turning

0.39

 Turn

0.39

 turned

0.38

 turning

0.35

 TURN

0.34

0.32

Turn

0.32

the

0.32

 कार

0.31

Activations Density 0.127%

instances of directional or positional movements and actions

No Comments

No Known Activations