INDEX

Explanations

references to the term 'Toy'

oai_token-act-pair · gpt-3.5-turbo

references to the "Toy Story" franchise and related elements

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

aunder

-0.73

terday

-0.72

xual

-0.72

icter

-0.70

 livest

-0.70

etheless

-0.68

idency

-0.67

 Hurricanes

-0.66

 untreated

-0.65

andestine

-0.65

POSITIVE LOGITS

ota

1.14

shop

0.87

hammer

0.87

Toy

0.85

owa

0.81

Chest

0.81

oga

0.80

ograp

0.80

 Crate

0.79

Toy

0.78

Activations Density 0.009%

references to the term 'Toy'

references to the "Toy Story" franchise and related elements

No Comments

No Known Activations