INDEX

Explanations

instances of the word "see."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 0-gemmascope-mlp-65k

Configuration

google/gemma-scope-2b-pt-mlp/layer_0/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.0.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

 tartalomajánló

-0.80

 WaitForSeconds

-0.63

RUnlock

-0.62

RectangleBorder

-0.61

olkata

-0.61

ieteur

-0.59

toMatchSnapshot

-0.59

 BoxDecoration

-0.59

 NSCoder

-0.58

see

-0.57

POSITIVE LOGITS

 ostavi

0.68

zzá

0.53

 насеље

0.52

Rated

0.52

 uved

0.50

 rated

0.49

RegressionTest

0.48

 hemsida

0.48

databind

0.48

 Rated

0.47

Activations Density 0.018%

instances of the word "see."

No Comments

No Known Activations