INDEX

Explanations

references to specific entities or nouns, such as organizations, people, or products

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

terday

-0.83

ĸļ

-0.76

oÄŁ

-0.74

 retrospect

-0.72

htt

-0.72

puter

-0.71

 derail

-0.70

 craw

-0.69

nsics

-0.68

 Conquest

-0.67

POSITIVE LOGITS

reet

1.37

reetings

1.29

ossip

1.28

oliath

1.27

asp

1.26

ourmet

1.26

irlfriend

1.24

CHQ

1.24

opher

1.23

ospels

1.23

Activations Density 8.975%

references to specific entities or nouns, such as organizations, people, or products

No Comments

No Known Activations