INDEX

Explanations

images with high resolution or specific details and labels related to people and events

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ormal

-0.98

cible

-0.90

gran

-0.90

ĪĴ

-0.83

rous

-0.83

adal

-0.81

soType

-0.80

rette

-0.79

gone

-0.78

ced

-0.78

POSITIVE LOGITS

 Images

1.33

 IMAGES

1.06

ystem

1.06

 Entertainment

1.05

Ltd

1.01

sburg

1.00

 caption

0.96

Library

0.96

 Library

0.94

 Agency

0.92

Activations Density 8.817%

images with high resolution or specific details and labels related to people and events

No Comments

No Known Activations