INDEX

Explanations

numeric expressions or codes depicted with symbols like '+' followed by numbers

oai_token-act-pair · gpt-3.5-turbo

numeric references and symbols indicating positive associations or ratings

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 Turing

-0.73

anium

-0.69

 Lord

-0.68

 Reloaded

-0.66

xton

-0.65

 Petty

-0.64

 artif

-0.57

 Sentinel

-0.56

 Progressive

-0.56

ãĥĪ

-0.56

POSITIVE LOGITS

/-

1.23

-+

0.87

/+

0.86

hot

0.81

-+-+-+-+

0.80

----------

0.79

Priv

0.79

xml

0.78

json

0.77

aign

0.77

Activations Density 0.020%

numeric expressions or codes depicted with symbols like '+' followed by numbers

numeric references and symbols indicating positive associations or ratings

No Comments

No Known Activations