INDEX

Explanations

words related to consequences or implications

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate the meaning or implications of a statement

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

thumbnails

-0.74

oked

-0.72

EStreamFrame

-0.68

oos

-0.67

uner

-0.65

Newsletter

-0.65

cart

-0.64

Kings

-0.63

taboola

-0.62

mens

-0.61

POSITIVE LOGITS

terday

1.03

hift

0.85

 goodbye

0.72

ãĥĨãĤ£

0.66

 è£ıè

0.66

Ðº

0.65

 Â±

0.65

 Lans

0.64

ãĤ¯

0.64

 passers

0.63

Activations Density 0.036%

words related to consequences or implications

phrases that indicate the meaning or implications of a statement

No Comments

No Known Activations