INDEX

Explanations

specific names or references to individuals

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Configuration

neuronpedia/gpt2-small__res_scl-ajt/6-res_scl-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

 tides

-0.24

iP

-0.24

 Mega

-0.24

Ott

-0.23

 Meier

-0.23

 Gomez

-0.23

 rape

-0.23

$$$$

-0.23

 cent

-0.22

 Gund

-0.22

POSITIVE LOGITS

manship

0.40

aws

0.36

wagen

0.34

hooting

0.34

alon

0.33

mallow

0.33

hift

0.32

sembly

0.32

luster

0.32

aber

0.31

Activations Density 7.331%

specific names or references to individuals

No Comments

No Known Activations