INDEX

Explanations

phrases related to authority figures making public statements

oai_token-act-pair · gpt-3.5-turbo

references to political positions or statements made publicly

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 unpop

-0.70

asionally

-0.57

 predec

-0.56

choes

-0.54

eteenth

-0.53

ommod

-0.53

Peb

-0.52

 heterogeneity

-0.50

 longstanding

-0.49

zens

-0.49

POSITIVE LOGITS

,,,,

0.94

to

0.87

"""

0.74

 unto

0.74

 towards

0.71

[/

0.69

[/

0.69

 thats

0.69

!!!!

0.67

""

0.67

Activations Density 0.780%

phrases related to authority figures making public statements

references to political positions or statements made publicly

No Comments

No Known Activations