INDEX

Explanations

phrases indicating speech or statements made by someone

oai_token-act-pair · gpt-3.5-turbo

instances of the word "said."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Lens

-0.68

ãĥĥãĥĪ

-0.68

ranean

-0.66

estern

-0.65

otin

-0.65

scan

-0.64

Pont

-0.63

email

-0.63

AMY

-0.63

à¦

-0.63

POSITIVE LOGITS

 goodbye

1.12

 bluntly

0.98

 sarcast

0.97

 aloud

0.96

 afterward

0.86

 unequivocally

0.78

 emphatically

0.77

 apologized

0.76

 repeatedly

0.76

:"

0.76

Activations Density 0.216%

phrases indicating speech or statements made by someone

instances of the word "said."

No Comments

No Known Activations