INDEX

Explanations

references to a specific publication, "The Toronto Star"

oai_token-act-pair · gpt-3.5-turbo

mentions of the term "Star."

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

Downloadha

-0.91

ĸļ

-0.84

ipop

-0.82

sembly

-0.81

ongyang

-0.80

ĵĺ

-0.78

ecause

-0.78

odcast

-0.78

terday

-0.76

essee

-0.75

POSITIVE LOGITS

vation

1.06

bucks

0.95

ved

0.95

light

0.92

ring

0.88

burst

0.86

ving

0.84

fish

0.83

buck

0.83

lings

0.78

Activations Density 0.022%

references to a specific publication, "The Toronto Star"

mentions of the term "Star."

No Comments

No Known Activations