INDEX

Explanations

any instances of the letter "P" in the text

oai_token-act-pair · gpt-3.5-turbo

the letter "P" or related sequences in the text

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

 Rebels

-0.88

 GOODMAN

-0.76

 diplom

-0.76

ģĸ

-0.71

 Orche

-0.69

 Breaking

-0.68

 Exodus

-0.66

adm

-0.64

 EntityItem

-0.63

ãĤ´ãĥ³

-0.62

POSITIVE LOGITS

aired

1.26

ardon

1.24

airs

1.18

ivot

1.18

ierce

1.15

ainted

1.14

appy

1.13

ussy

1.11

adding

1.10

uls

1.07

Activations Density 0.041%

any instances of the letter "P" in the text

the letter "P" or related sequences in the text

No Comments

No Known Activations