INDEX

Explanations

references to human anatomy, specifically arms

oai_token-act-pair · gpt-3.5-turbo

references to weapons or the arms industry

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 Woodward

-0.84

 Dough

-0.77

ãĤ©

-0.72

TION

-0.70

ded

-0.70

advertisement

-0.68

IENCE

-0.66

 Sound

-0.65

 Vide

-0.65

gres

-0.64

POSITIVE LOGITS

ament

1.14

aments

0.94

 arms

0.90

chair

0.89

ageddon

0.86

arm

0.81

illary

0.80

guards

0.79

bands

0.77

phabet

0.76

Activations Density 0.009%

references to human anatomy, specifically arms

references to weapons or the arms industry

No Comments

No Known Activations