INDEX

Explanations

words related to political figures and controversies

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

mentions of 'L.'

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 12-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_post

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_post

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

 Ceres

-0.66

QC

-0.63

 Constantin

-0.63

 Satanic

-0.61

 plum

-0.60

 Guan

-0.60

coy

-0.59

 head

-0.58

 Mend

-0.57

 folk

-0.57

POSITIVE LOGITS

aundering

1.54

osing

1.37

aughed

1.32

ooting

1.32

iking

1.32

aughs

1.30

oneliness

1.27

obb

1.26

obbies

1.25

apses

1.24

Activations Density 0.025%

words related to political figures and controversies

mentions of 'L.'

No Comments

No Known Activations