INDEX

Explanations

proper nouns or specific entities, probably related to a type of classification or identification

oai_token-act-pair · gpt-3.5-turbo

the term "Other" in various contexts

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

ardo

-0.64

 fres

-0.63

Â´

-0.62

usted

-0.60

 daring

-0.58

cour

-0.56

itude

-0.56

lif

-0.54

 nearly

-0.54

zos

-0.54

POSITIVE LOGITS

 Other

3.37

 Others

2.59

Other

2.37

Others

1.99

 OTHER

1.96

other

1.72

 Another

1.59

 other

1.54

 Various

1.49

OTHER

1.34

Activations Density 0.007%

proper nouns or specific entities, probably related to a type of classification or identification

the term "Other" in various contexts

No Comments

No Known Activations