INDEX

Explanations

phrases addressing or referring to the person being spoken to

oai_token-act-pair · gpt-3.5-turbo

references to the second person

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

eca

-0.79

GMT

-0.78

emon

-0.78

ĺħ

-0.77

ĸļ

-0.72

inction

-0.71

ģ«

-0.71

¿½

-0.70

entimes

-0.70

asio

-0.69

POSITIVE LOGITS

 guys

1.32

're

1.17

 yourselves

1.08

tub

0.94

'll

0.86

 gentlemen

0.85

sir

0.84

 filthy

0.83

've

0.82

 majesty

0.80

Activations Density 0.196%

phrases addressing or referring to the person being spoken to

references to the second person

No Comments

No Known Activations