INDEX

Explanations

words related to professional titles or positions

oai_token-act-pair · gpt-3.5-turbo

words related to individuals or entities in positions of authority or influence

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 fresh

-0.67

 vain

-0.65

 stamp

-0.64

 Stamp

-0.60

 Haley

-0.60

 damn

-0.59

Pwr

-0.59

 Ange

-0.59

EStream

-0.59

 fist

-0.58

POSITIVE LOGITS

ors

4.32

ORS

2.80

or

2.01

ours

1.97

orship

1.92

ored

1.82

oring

1.78

OR

1.73

orate

1.62

orm

1.55

Activations Density 0.014%

words related to professional titles or positions

words related to individuals or entities in positions of authority or influence

No Comments

No Known Activations