INDEX

Explanations

phrases related to destruction and criminal activities

oai_token-act-pair · gpt-3.5-turbo

references to data systems and physical infrastructure being impacted or damaged

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ITNESS

-0.72

rosso

-0.65

oneliness

-0.63

algia

-0.63

annot

-0.62

irection

-0.61

iasis

-0.60

apest

-0.59

METHOD

-0.59

onomous

-0.59

POSITIVE LOGITS

 indiscrim

0.88

 causing

0.87

 injuring

0.83

 havoc

0.81

 violently

0.80

 prematurely

0.79

 alike

0.78

destroy

0.75

 momentarily

0.73

 belonging

0.73

Activations Density 0.416%

phrases related to destruction and criminal activities

references to data systems and physical infrastructure being impacted or damaged

No Comments

No Known Activations