INDEX

Explanations

mentions of hate crimes and related incidents involving discrimination and violence

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

æ©Ł

-1.13

clinton

-1.05

aver

-1.05

interstitial

-1.03

ODE

-0.99

å§«

-0.99

 reluct

-0.97

debian

-0.94

é¾įå

-0.93

amina

-0.92

POSITIVE LOGITS

fully

1.80

fulness

1.54

ful

1.43

 vengeance

1.14

 crimes

1.11

FUL

1.07

 hate

1.01

 revenge

0.99

lihood

0.98

lessly

0.96

Activations Density 0.988%

mentions of hate crimes and related incidents involving discrimination and violence

No Comments

No Known Activations