INDEX

Explanations

instances of the word "pants" in the text

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "pants"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

Reviewed

-0.74

 Flavoring

-0.67

=-=-=-=-

-0.67

 Galile

-0.66

AUT

-0.66

Fund

-0.64

 Democr

-0.64

SPONSORED

-0.64

 Canaver

-0.64

ILY

-0.63

POSITIVE LOGITS

 pants

1.19

 trousers

1.16

pants

1.14

uits

1.03

bag

1.00

leeve

1.00

belt

0.98

uit

0.95

 jeans

0.94

shirt

0.94

Activations Density 0.009%

instances of the word "pants" in the text

occurrences of the word "pants"

No Comments

No Known Activations