INDEX

Explanations

prepositions

oai_token-act-pair · gpt-3.5-turbo

phrases indicating transition or transformation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

hair

-0.70

 IMAGES

-0.62

 Cookies

-0.60

lli

-0.59

Cod

-0.59

Dream

-0.59

ice

-0.59

worms

-0.57

PI

-0.57

 Dise

-0.56

POSITIVE LOGITS

 downright

0.80

 becoming

0.80

 outright

0.80

 adulthood

0.75

 fully

0.74

 realizing

0.73

 embracing

0.72

 suddenly

0.71

 solidly

0.69

adier

0.69

Activations Density 0.148%

prepositions

phrases indicating transition or transformation

No Comments

No Known Activations