INDEX

Explanations

references to intelligence agencies

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Configuration

neuronpedia/gpt2-small__res_scefr-ajt/6-res_scefr-ajt

Prompts (Dashboard)

12,288 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

46,080

Data Type

torch.float32

Hook Point

blocks.6.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

apollo-research/Skylion007-openwebtext-tokenizer-gpt2

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

mans

-1.22

days

-1.03

flight

-1.03

bilt

-1.00

manship

-1.00

ifiers

-0.97

mann

-0.96

speed

-0.96

language

-0.95

metic

-0.95

POSITIVE LOGITS

KE

1.62

AMI

1.58

PS

1.49

HAEL

1.42

KER

1.38

YA

1.36

OSH

1.27

XT

1.24

RE

1.24

LE

1.23

Activations Density 0.447%

references to intelligence agencies

No Comments

No Known Activations