INDEX

Explanations

the word "cabin"

oai_token-act-pair · gpt-3.5-turbo

references to "cabin" and its variations

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

cz

-0.78

llular

-0.76

pace

-0.73

terday

-0.70

nesota

-0.68

posing

-0.65

Ô

-0.65

FW

-0.63

lda

-0.63

mington

-0.63

POSITIVE LOGITS

etry

0.96

oodle

0.89

illo

0.88

aughs

0.87

rio

0.74

oline

0.74

yrinth

0.73

aret

0.72

acle

0.70

 cabin

0.68

Activations Density 0.087%

the word "cabin"

references to "cabin" and its variations

No Comments

No Known Activations