INDEX

Explanations

phrases related to controversy, conflict, and unwanted consequences

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

meric

-0.78

anson

-0.74

eston

-0.70

ials

-0.67

este

-0.67

sembly

-0.66

sie

-0.63

hi

-0.62

edy

-0.61

abor

-0.61

POSITIVE LOGITS

 hotter

0.63

flies

0.63

ãĤ¤ãĥĪ

0.62

 missiles

0.61

lda

0.61

ULTS

0.61

izons

0.60

 exting

0.60

fires

0.60

fire

0.60

Activations Density 10.088%

phrases related to controversy, conflict, and unwanted consequences

No Comments

No Known Activations