INDEX

Explanations

names of individuals, particularly those with the word 'ank' in them

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

 Scotia

-1.29

KT

-1.10

pha

-1.03

pert

-1.03

abetic

-0.94

 scattering

-0.92

RFC

-0.91

ãĥ¤

-0.91

idia

-0.89

 Glob

-0.85

POSITIVE LOGITS

er

1.41

ernel

1.30

erman

1.29

erous

1.27

owski

1.27

ers

1.25

erness

1.23

erson

1.22

ety

1.22

ular

1.21

Activations Density 2.001%

names of individuals, particularly those with the word 'ank' in them

No Comments

No Known Activations