INDEX

Explanations

phrases related to military operations and security measures

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

fair

-0.65

former

-0.65

hops

-0.64

sburg

-0.63

late

-0.63

chan

-0.63

press

-0.63

sov

-0.62

NN

-0.59

ouf

-0.59

POSITIVE LOGITS

stood

1.61

 regards

1.59

 regard

1.54

 impunity

1.47

draw

1.43

drawn

1.38

 respect

1.30

standing

1.26

holding

1.15

 dignity

1.03

Activations Density 0.655%

phrases related to military operations and security measures

No Comments

No Known Activations