INDEX

Explanations

parenthesis with numbers indicating emphasis

oai_token-act-pair · gpt-3.5-turbo

opening parentheses used in expressions or statements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 wildlife

-0.76

 retard

-0.73

 glare

-0.70

 clin

-0.70

 lull

-0.69

 flow

-0.69

 incons

-0.69

 millenn

-0.69

 daily

-0.66

 inund

-0.65

POSITIVE LOGITS

although

1.59

albeit

1.59

including

1.56

which

1.51

except

1.49

especially

1.48

possibly

1.48

perhaps

1.47

see

1.46

usually

1.45

Activations Density 0.155%

parenthesis with numbers indicating emphasis

opening parentheses used in expressions or statements

No Comments

No Known Activations