INDEX

Explanations

words related to communication, information dissemination, and opinions

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

addafi

-0.91

zzi

-0.83

 Wrestle

-0.82

 Breaker

-0.82

 Mata

-0.82

atari

-0.80

 Able

-0.78

 Columb

-0.78

 McAuliffe

-0.77

 Clemson

-0.76

POSITIVE LOGITS

manship

1.34

emies

1.31

terday

1.23

umer

1.21

kowski

1.21

burgh

1.19

kamp

1.17

ource

1.15

chen

1.13

chant

1.11

Activations Density 0.586%

words related to communication, information dissemination, and opinions

No Comments

No Known Activations