INDEX

Explanations

verbs related to emotional responses and political actions

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 intel

-0.91

 dust

-0.86

 exit

-0.85

 managers

-0.85

 meeting

-0.84

 actors

-0.84

 sweat

-0.84

 working

-0.83

 Europa

-0.81

 directors

-0.80

POSITIVE LOGITS

Ĭ

1.42

ŀ

1.40

uing

1.37

ety

1.33

¼

1.31

ª

1.29

´

1.29

ues

1.29

ĳ

1.27

ued

1.27

Activations Density 2.211%

verbs related to emotional responses and political actions

No Comments

No Known Activations