INDEX

Explanations

proper nouns indicating public figures or organizations

oai_token-act-pair · gpt-3.5-turbo

instances of the word "called" indicating references to statements or actions made by individuals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

inth

-0.87

 istg

-0.74

bilt

-0.71

Cub

-0.69

 unden

-0.65

 bunny

-0.64

 Column

-0.64

abba

-0.64

bourne

-0.63

 guiName

-0.62

POSITIVE LOGITS

call

1.03

calling

0.95

 calling

0.85

 Call

0.83

 called

0.82

 Calling

0.81

Call

0.80

 upon

0.79

0.77

backs

0.75

Activations Density 0.056%

proper nouns indicating public figures or organizations

instances of the word "called" indicating references to statements or actions made by individuals

No Comments

No Known Activations