INDEX

Explanations

references to political positions, particularly related to governmental roles or controversies

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

ISM

-0.77

 Lions

-0.76

 Abyssal

-0.71

 Salvation

-0.71

 brackets

-0.69

 Worlds

-0.67

STAR

-0.66

 dots

-0.66

ritch

-0.66

 Italians

-0.66

POSITIVE LOGITS

chool

1.31

cription

1.29

ervation

1.21

umed

1.20

erves

1.19

byter

1.17

ervative

1.15

erving

1.11

idency

1.11

erver

1.10

Activations Density 6.676%

references to political positions, particularly related to governmental roles or controversies

No Comments

No Known Activations