INDEX

Explanations

words related to political figure "Abbott"

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 Hiroshima

-0.79

lihood

-0.74

 Cherokee

-0.68

hare

-0.66

 Siberian

-0.65

pace

-0.64

arning

-0.64

boarding

-0.63

 Chero

-0.63

DonaldTrump

-0.62

POSITIVE LOGITS

abad

1.16

 Turnbull

0.85

aby

0.82

shire

0.76

 Abbott

0.76

 Laboratories

0.73

onom

0.71

DEF

0.71

elling

0.70

minster

0.69

Activations Density 0.034%

words related to political figure "Abbott"

No Comments

No Known Activations