INDEX

Explanations

references to body parts, especially arms

oai_token-act-pair · gpt-3.5-turbo

mentions of "arm" in various contexts

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Sound

-0.79

 Vide

-0.76

 Dough

-0.70

BUR

-0.67

 Atmosp

-0.66

ween

-0.65

VEN

-0.64

iculty

-0.64

 Rewards

-0.63

 Judicial

-0.63

POSITIVE LOGITS

ageddon

1.33

aments

1.32

chair

1.32

ament

1.30

illary

1.05

strap

1.02

chairs

1.00

arm

0.99

ength

0.95

amental

0.91

Activations Density 0.013%

references to body parts, especially arms

mentions of "arm" in various contexts

No Comments

No Known Activations