INDEX

Explanations

mentions of lethal or dangerous items, actions, or situations

oai_token-act-pair · gpt-3.5-turbo

references to lethal substances or actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

oard

-0.83

ilage

-0.81

aliation

-0.76

resses

-0.75

bed

-0.72

æĦ

-0.72

roller

-0.72

roll

-0.72

baby

-0.71

acular

-0.69

POSITIVE LOGITS

cious

0.94

istics

0.93

uania

0.84

cius

0.80

lihood

0.75

SPONSORED

0.74

 Luthor

0.74

istically

0.73

 injection

0.72

 ANGEL

0.71

Activations Density 0.035%

mentions of lethal or dangerous items, actions, or situations

references to lethal substances or actions

No Comments

No Known Activations