INDEX

Explanations

keyword "genocide."

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "genocide."

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

amina

-0.83

PATH

-0.75

Lear

-0.72

Higher

-0.71

aton

-0.69

alg

-0.68

Lago

-0.67

erm

-0.67

upper

-0.67

 Eucl

-0.66

POSITIVE LOGITS

 genocide

1.40

 Genocide

1.21

ocide

1.14

 massacres

0.95

 atrocities

0.91

ãĥ¼ãĥĨãĤ£

0.90

 massac

0.89

ãĤº

0.88

 massacre

0.83

 extermination

0.82

Activations Density 0.017%

keyword "genocide."

occurrences of the word "genocide."

No Comments

No Known Activations