INDEX

Explanations

phrases or sentences prompting the reader to use their imagination

oai_token-act-pair · gpt-3.5-turbo

instances of the phrase "can you imagine."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

bard

-0.74

inals

-0.73

ldon

-0.69

cific

-0.68

ged

-0.66

wa

-0.62

die

-0.62

fund

-0.60

anski

-0.59

church

-0.58

POSITIVE LOGITS

ĸļ

0.88

lihood

0.82

 aloud

0.76

RON

0.75

 imagine

0.74

âĶĢâĶĢâĶĢâĶĢ

0.73

orial

0.72

how

0.68

peak

0.67

 Â¯

0.65

Activations Density 0.022%

phrases or sentences prompting the reader to use their imagination

instances of the phrase "can you imagine."

No Comments

No Known Activations