INDEX

Explanations

political and leadership-related terms and names

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

Magikarp

-0.99

illac

-0.97

lus

-0.91

afort

-0.91

hammad

-0.89

inventoryQuantity

-0.87

%%%%

-0.86

nces

-0.84

fw

-0.84

 Ukrain

-0.83

POSITIVE LOGITS

 Corbyn

1.50

URN

0.90

uckle

0.90

nuts

0.87

urger

0.86

 Peaks

0.86

shire

0.86

 Hots

0.83

IGHTS

0.83

 Clarkson

0.82

Activations Density 1.314%

political and leadership-related terms and names

No Comments

No Known Activations