INDEX

Explanations

phrases related to written or verbal descriptions of various items or situations

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

nuts

-0.79

avis

-0.68

igham

-0.66

yrus

-0.65

abb

-0.65

 scrimmage

-0.65

BSD

-0.65

ced

-0.63

aghetti

-0.63

uld

-0.62

POSITIVE LOGITS

 description

0.92

 descriptions

0.91

anguage

0.84

 synopsis

0.79

REDACTED

0.78

description

0.77

terday

0.76

 describ

0.75

 thereof

0.75

>[

0.74

Activations Density 0.020%

phrases related to written or verbal descriptions of various items or situations

No Comments

No Known Activations