INDEX

Explanations

references to specific authors or sources at the beginning of text excerpts

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate authorship or attribution in texts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 grip

-0.63

 fabric

-0.61

 standing

-0.59

 listed

-0.58

 current

-0.58

 right

-0.57

 muff

-0.56

 hood

-0.56

 origin

-0.55

 green

-0.55

POSITIVE LOGITS

By

3.25

By

1.95

by

1.83

BY

1.79

Through

1.67

Around

1.48

From

1.41

ccording

1.40

With

1.40

Within

1.39

Activations Density 0.014%

references to specific authors or sources at the beginning of text excerpts

phrases that indicate authorship or attribution in texts

No Comments

No Known Activations