INDEX

Explanations

words expressing personal opinions or perspectives

oai_token-act-pair · gpt-3.5-turbo

occurrences of the pronoun "I"

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

 otherwise

-0.66

 sustain

-0.64

ens

-0.62

 equals

-0.58

 equal

-0.57

 aver

-0.56

equ

-0.56

his

-0.55

 apes

-0.54

 established

-0.54

POSITIVE LOGITS

2.97

My

1.73

IAS

1.53

Is

1.43

IU

1.40

Honestly

1.39

II

1.38

1.36

We

1.35

Personally

1.34

Activations Density 0.076%

words expressing personal opinions or perspectives

occurrences of the pronoun "I"

No Comments

No Known Activations