INDEX

Explanations

phrases related to actions and intentions of people

oai_token-act-pair · gpt-3.5-turbo

references to individuals and their actions or claims

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

Flight

-0.69

artifacts

-0.63

taboola

-0.63

endif

-0.61

gettable

-0.61

NX

-0.61

ahoo

-0.60

else

-0.60

Fine

-0.59

ival

-0.58

POSITIVE LOGITS

 championed

1.02

 profess

0.96

 purportedly

0.95

 supposedly

0.91

esp

0.91

 ostensibly

0.91

 purported

0.88

 preached

0.84

 nurt

0.82

 preach

0.81

Activations Density 0.206%

phrases related to actions and intentions of people

references to individuals and their actions or claims

No Comments

No Known Activations