INDEX

Explanations

phrases related to violence or conflict, especially those involving blood

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.8.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

awaru

-0.88

 Amend

-0.84

IX

-0.75

 Lank

-0.72

OPLE

-0.70

merce

-0.70

VICE

-0.68

acle

-0.68

ECH

-0.67

 Spac

-0.66

POSITIVE LOGITS

thirst

1.49

hound

1.29

bath

1.28

stained

1.28

lust

1.18

shed

1.08

lines

1.00

spl

0.92

 stains

0.92

shot

0.91

Activations Density 0.023%

phrases related to violence or conflict, especially those involving blood

No Comments

No Known Activations