INDEX

Explanations

the word "go" followed by a number that represents an action or location

oai_token-act-pair · gpt-3.5-turbo

instances of the word "go"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

ullah

-0.74

ament

-0.72

ificent

-0.68

itionally

-0.68

oret

-0.66

 Horus

-0.63

icio

-0.62

uctor

-0.62

oral

-0.61

inguished

-0.60

POSITIVE LOGITS

vt

1.14

verning

1.02

lems

1.00

ggle

0.98

 overboard

0.88

Ń·

0.83

etz

0.83

 ahead

0.83

aded

0.82

ALK

0.82

Activations Density 0.095%

the word "go" followed by a number that represents an action or location

instances of the word "go"

No Comments

No Known Activations