INDEX

Explanations

proper nouns or names of organizations

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

ä»£

-0.31

Hos

-0.28

EMENT

-0.28

ISTER

-0.28

 Presbyterian

-0.28

UN

-0.27

IRO

-0.27

çĶ°

-0.26

TOR

-0.26

Ms

-0.26

POSITIVE LOGITS

 nost

0.29

idal

0.29

 tail

0.28

¥ŀ

0.27

 unden

0.27

 undermin

0.27

 booze

0.27

 manif

0.26

 imperson

0.26

phr

0.26

Activations Density 0.086%

proper nouns or names of organizations

No Comments

No Known Activations