INDEX

Explanations

groups or organizations involved in conflict or military operations

oai_token-act-pair · gpt-3.5-turbo

references to specific militant groups and individuals involved in conflicts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

RIP

-0.78

 Ellis

-0.69

 Scotch

-0.68

 Eucl

-0.68

ym

-0.67

Scott

-0.67

Ven

-0.67

PASS

-0.66

EB

-0.66

GST

-0.63

POSITIVE LOGITS

usra

1.37

Nusra

1.28

 Palestin

0.94

ascus

0.93

prus

0.86

 eleph

0.83

Syrian

0.81

 guiActiveUn

0.81

VIDIA

0.79

 conflic

0.78

Activations Density 0.010%

groups or organizations involved in conflict or military operations

references to specific militant groups and individuals involved in conflicts

No Comments

No Known Activations