INDEX

Explanations

proper nouns or people's names in relation to actions or events

oai_token-act-pair · gpt-3.5-turbo

phrases indicating actions taken against an individual or character judgments about them

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.9.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

zbollah

-0.86

anguage

-0.65

ses

-0.58

£ı

-0.53

okia

-0.53

opoly

-0.53

 neigh

-0.53

 teasp

-0.52

 unpop

-0.52

pora

-0.52

POSITIVE LOGITS

him

2.27

him

1.64

HIM

1.46

Him

1.28

He

0.83

his

0.83

essor

0.81

His

0.80

he

0.79

his

0.79

Activations Density 1.153%

proper nouns or people's names in relation to actions or events

phrases indicating actions taken against an individual or character judgments about them

No Comments

No Known Activations