INDEX

Explanations

mentions of the word "brain"

oai_token-act-pair · gpt-3.5-turbo

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

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

IFrame

Link

Negative Logits

 Bundy

-0.73

nesday

-0.70

 FANTASY

-0.69

 Arabia

-0.66

 Bowie

-0.66

adoes

-0.65

 impunity

-0.65

 Faust

-0.63

Mex

-0.62

Riy

-0.62

POSITIVE LOGITS

stem

1.38

washed

1.26

washing

1.18

wash

1.11

iac

1.07

waves

0.99

fuck

0.92

dead

0.89

storms

0.89

caps

0.87

Activations Density 0.019%

mentions of the word "brain"

No Known Activations