INDEX

Explanations

phrases related to professional or formal contexts

oai_token-act-pair · gpt-3.5-turbo

punctuation marks and their related contexts within the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 engagements

-0.72

lections

-0.65

 attent

-0.63

 engagement

-0.62

aeda

-0.62

 sacrifice

-0.60

 equip

-0.60

 allegiance

-0.59

 sooner

-0.59

uno

-0.59

POSITIVE LOGITS

However

0.89

Anonymous

0.88

SEE

0.86

Newsletter

0.86

Article

0.85

Still

0.84

ccording

0.84

Scroll

0.81

Researchers

0.81

Prosecutors

0.81

Activations Density 0.510%

phrases related to professional or formal contexts

punctuation marks and their related contexts within the text

No Comments

No Known Activations