INDEX

Explanations

phrases related to small or miniature versions of objects or concepts

oai_token-act-pair · gpt-3.5-turbo

instances of the word "mini" across various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

âķĲâķĲ

-0.83

Ö¼

-0.76

xual

-0.72

rer

-0.66

yer

-0.65

ILLE

-0.65

KI

-0.65

mingham

-0.64

 defe

-0.63

worthiness

-0.62

POSITIVE LOGITS

ature

1.32

atur

1.28

atures

1.02

ocular

0.90

 sized

0.81

oby

0.79

eco

0.78

ota

0.78

atomic

0.76

istered

0.75

Activations Density 0.013%

phrases related to small or miniature versions of objects or concepts

instances of the word "mini" across various contexts

No Comments

No Known Activations