INDEX

Explanations

mentions of clothing items, particularly pants

oai_token-act-pair · gpt-3.5-turbo

references to pants or related clothing items

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 Scotia

-0.74

AUT

-0.71

 Galile

-0.70

 Flavoring

-0.69

KNOWN

-0.66

=-=-=-=-

-0.66

ormon

-0.64

 Woodward

-0.64

Reviewed

-0.63

 uncertainties

-0.63

POSITIVE LOGITS

 pants

1.12

 trousers

1.07

uits

1.05

pants

1.01

leeve

1.00

uit

1.00

bag

1.00

belt

0.97

face

0.94

 jeans

0.91

Activations Density 0.022%

mentions of clothing items, particularly pants

references to pants or related clothing items

No Comments

No Known Activations