INDEX

Explanations

phrases related to professional decision-making and discussion

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

FTWARE

-0.70

ilial

-0.66

axy

-0.65

azeera

-0.63

akia

-0.62

âĶĢâĶĢâĶĢâĶĢâĶĢâĶĢâĶĢâĶĢ

-0.62

iliar

-0.62

STAR

-0.60

holm

-0.60

isible

-0.59

POSITIVE LOGITS

fy

0.88

 somebody

0.80

 someone

0.80

 they

0.71

yip

0.70

soever

0.69

 anyone

0.68

 anybody

0.67

 anything

0.66

someone

0.63

Activations Density 8.826%

phrases related to professional decision-making and discussion

No Comments

No Known Activations