INDEX

Explanations

variations of the word "scale" in different contexts

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

Pom

-0.40

 Hoffman

-0.38

enumi

-0.38

 Quaternion

-0.38

 PyObject

-0.38

%@

-0.37

DJ

-0.37

Ợ

-0.37

WidgetItem

-0.37

Pub

-0.36

POSITIVE LOGITS

 scale

4.22

 Scale

3.95

Scale

3.66

scale

3.64

 SCALE

3.55

 scales

3.53

SCALE

3.13

 Scales

3.08

 scaling

3.06

 scaled

2.98

Activations Density 0.169%

variations of the word "scale" in different contexts

No Comments

No Known Activations