INDEX

Explanations

references related to popular culture figures and TV show characters, particularly focusing on well-known names and activities

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

umar

-0.96

oran

-0.90

 ranc

-0.85

rongh

-0.81

sth

-0.79

 Milton

-0.78

vironment

-0.77

ionage

-0.77

ilk

-0.76

hoe

-0.75

POSITIVE LOGITS

 pageant

1.25

Girls

1.21

 Contest

1.15

 contestants

1.14

 endors

1.05

 actresses

1.05

 cosmetics

1.04

 contestant

1.04

models

1.04

rities

0.98

Activations Density 3.200%

references related to popular culture figures and TV show characters, particularly focusing on well-known names and activities

No Comments

No Known Activations