INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-mlp-65k

Configuration

google/gemma-scope-2b-pt-mlp/layer_18/width_65k/average_l0_88

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.18.hook_mlp_out

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

InputBorder

-0.68

RectangleBorder

-0.62

IndentedString

-0.59

CodeAttribute

-0.57

ImageContext

-0.53

ContentAsync

-0.50

aarrggbb

-0.50

AddHtmlAttribute

-0.46

ipedi

-0.46

 testé

-0.45

POSITIVE LOGITS

########.

0.70

prnt

0.54

enumii

0.54

LLocation

0.53

codable

0.51

 Rien

0.51

comm

0.51

phi

0.50

berglass

0.50

 crossorigin

0.50

Activations Density 0.000%

No Known Activations

This feature has no known activations.