INDEX

Explanations

references related to historical events and military actions

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 Kardash

-0.98

 Transparency

-0.94

ï¸ı

-0.84

Kids

-0.82

creator

-0.80

 wellness

-0.78

 parenting

-0.77

 Consumer

-0.77

Consumer

-0.76

 Blockchain

-0.75

POSITIVE LOGITS

 garrison

1.59

 infantry

1.44

 artillery

1.40

 cavalry

1.37

 Allied

1.37

 troops

1.33

 besie

1.30

 Infantry

1.27

 battalion

1.26

 skirm

1.25

Activations Density 0.629%

references related to historical events and military actions

No Comments

No Known Activations