INDEX

Explanations

symbols and punctuation

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 13-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_13/width_65k/average_l0_75

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.13.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

 Monfieur

-1.02

featureID

-1.02

 AssemblyCulture

-1.00

setVerticalGroup

-1.00

 Normdatei

-0.96

tvguidetime

-0.96

Obrázky

-0.93

 iſt

-0.93

principalTable

-0.91

 vectorstock

-0.89

POSITIVE LOGITS

 still

0.56

0.50

0.49

0.48

往

0.47

as

0.46

 humor

0.46

 fully

0.45

Activations Density 0.021%

symbols and punctuation

No Known Activations