INDEX

Explanations

words related to involvement and interaction

oai_token-act-pair · gpt-3.5-turbo

instances of the word "engagement."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

\\\\\\\\

-0.69

olog

-0.68

 lethal

-0.65

 Sabha

-0.65

uggage

-0.65

uran

-0.64

printed

-0.63

mia

-0.63

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

-0.63

 surviving

-0.62

POSITIVE LOGITS

 engagement

1.07

agement

0.87

naire

0.83

 engaged

0.83

 engagements

0.78

ATURE

0.71

ernaut

0.70

 engages

0.69

EMENT

0.68

hips

0.68

Activations Density 0.013%

words related to involvement and interaction

instances of the word "engagement."

No Comments

No Known Activations