INDEX

Explanations

contractions of the word 'isn't'

oai_token-act-pair · gpt-3.5-turbo

repetitive use of the word "it" in rhetorical questions

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

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

iper

-0.59

digit

-0.58

elvet

-0.57

adra

-0.57

ERG

-0.56

ersed

-0.56

bard

-0.56

Ĥİ

-0.56

comings

-0.55

anders

-0.55

POSITIVE LOGITS

 thou

1.02

 they

0.89

it

0.87

 Thou

0.82

 everybody

0.82

you

0.80

we

0.79

 there

0.76

he

0.76

 anybody

0.75

Activations Density 0.052%

contractions of the word 'isn't'

repetitive use of the word "it" in rhetorical questions

No Comments

No Known Activations