INDEX

Explanations

references to military troops

oai_token-act-pair · gpt-3.5-turbo

mentions of military troops

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

Ã±

-0.70

izz

-0.67

ÃŁ

-0.67

bid

-0.67

 partName

-0.67

 Yelp

-0.65

 Ross

-0.65

gie

-0.64

velop

-0.63

 Bloom

-0.63

POSITIVE LOGITS

 troops

3.48

 soldiers

2.34

 troop

2.31

 armies

1.97

 troopers

1.93

 Soldiers

1.77

 marines

1.75

 generals

1.68

 forces

1.66

 Marines

1.60

Activations Density 0.013%

references to military troops

mentions of military troops

No Comments

No Known Activations