INDEX

Explanations

instances of the word "kit" in the text

oai_token-act-pair · gpt-3.5-turbo

references to kits and related products

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

arching

-0.87

unct

-0.80

 Observatory

-0.72

nesota

-0.68

acebook

-0.66

ews

-0.66

University

-0.62

rir

-0.61

unknown

-0.61

 Liberties

-0.60

POSITIVE LOGITS

kit

1.30

 kits

1.24

chens

1.21

kit

1.15

Kit

0.95

 Kits

0.88

kat

0.81

chy

0.76

 Kats

0.75

Kit

0.74

Activations Density 0.007%

instances of the word "kit" in the text

references to kits and related products

No Comments

No Known Activations