INDEX

Explanations

phrases related to military and political actions, particularly airstrikes and military interventions in specific regions

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

brow

-0.68

Consumer

-0.68

ãĤ¦ãĤ¹

-0.68

Tickets

-0.68

kW

-0.66

 Edison

-0.65

 partName

-0.64

Interstitial

-0.62

 Toast

-0.62

mith

-0.61

POSITIVE LOGITS

 Syria

1.38

 Iraq

1.35

 Syrian

1.32

 Libya

1.29

Syria

1.28

 Libyan

1.25

 Iraqi

1.23

Iraq

1.21

 Afghanistan

1.20

Syrian

1.20

Activations Density 1.203%

phrases related to military and political actions, particularly airstrikes and military interventions in specific regions

No Comments

No Known Activations