INDEX

Explanations

words related to guns, shootings, and violence

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

 Archdemon

-0.67

 Strawberry

-0.65

atta

-0.64

ç«

-0.64

 Hort

-0.64

gres

-0.63

annis

-0.63

akuya

-0.63

 Lemon

-0.61

Sat

-0.61

POSITIVE LOGITS

powder

1.45

 rifles

1.06

 handguns

1.06

 guns

1.03

 muzzle

1.02

smith

1.01

 pistols

1.00

guns

1.00

 ammunition

0.99

 rifle

0.99

Activations Density 2.311%

words related to guns, shootings, and violence

No Comments

No Known Activations