INDEX

Explanations

words related to email correspondences

oai_token-act-pair · gpt-3.5-turbo

instances of the word "Emails."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

own

-0.80

vol

-0.77

lust

-0.77

 Turner

-0.74

 Virgin

-0.74

ded

-0.69

ãĥĵ

-0.69

Stars

-0.67

 Raleigh

-0.67

 Durham

-0.67

POSITIVE LOGITS

 Emails

1.58

ileaks

0.99

catentry

0.92

achev

0.92

 inbox

0.85

ÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤ

0.85

 emails

0.82

 undermin

0.81

ogly

0.77

 correspondence

0.77

Activations Density 0.003%

words related to email correspondences

instances of the word "Emails."

No Comments

No Known Activations