INDEX

Explanations

the words related to originality, especially focusing on the term "Original" with a very high activation value

oai_token-act-pair · gpt-3.5-turbo

mentions of originality in creative works

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

wal

-0.75

walk

-0.72

ucket

-0.72

=-=-=-=-=-=-=-=-

-0.71

oned

-0.67

rolet

-0.67

ennett

-0.66

ega

-0.66

washer

-0.65

 Simulator

-0.64

POSITIVE LOGITS

ity

1.26

ITY

1.07

izations

0.90

ities

0.81

itÃ©

0.80

lly

0.75

itiz

0.74

itized

0.74

ITIES

0.73

smanship

0.73

Activations Density 0.017%

the words related to originality, especially focusing on the term "Original" with a very high activation value

mentions of originality in creative works

No Comments

No Known Activations