INDEX

Explanations

instances where an individual is taking credit for something or acknowledging their own actions

oai_token-act-pair · gpt-3.5-turbo

the phrase "I have" 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 @ 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

oshi

-0.72

 Apart

-0.65

catentry

-0.64

TG

-0.63

earable

-0.60

icking

-0.59

bies

-0.59

ourgeois

-0.57

patch

-0.56

weed

-0.56

POSITIVE LOGITS

 been

1.28

been

1.08

 undergone

0.92

 seen

0.90

 gotten

0.89

 become

0.88

 spoken

0.87

 Been

0.87

 begun

0.86

 learnt

0.82

Activations Density 0.263%

instances where an individual is taking credit for something or acknowledging their own actions

the phrase "I have" in various contexts

No Comments

No Known Activations