INDEX

Explanations

phrases or terms related to possessing a significant or symbolic role

oai_token-act-pair · gpt-3.5-turbo

phrases indicating authority or representation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

uyomi

-0.79

asing

-0.77

>(

-0.68

kat

-0.68

kward

-0.66

 WATCHED

-0.65

externalActionCode

-0.65

MRI

-0.64

jud

-0.64

fur

-0.64

POSITIVE LOGITS

 sorts

1.21

 excellence

0.95

 choice

0.93

 mankind

0.90

 civilization

0.89

 humankind

0.87

 sanity

0.84

 civilisation

0.81

 multicultural

0.79

 tomorrow

0.79

Activations Density 0.194%

phrases or terms related to possessing a significant or symbolic role

phrases indicating authority or representation

No Comments

No Known Activations