INDEX

Explanations

keywords related to wars and conflicts

oai_token-act-pair · gpt-3.5-turbo

references to war

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

aha

-0.75

enna

-0.68

icken

-0.67

eca

-0.64

aque

-0.64

 deposit

-0.63

ess

-0.61

iae

-0.60

atin

-0.59

 authorities

-0.59

POSITIVE LOGITS

WAR

3.89

WAR

2.69

War

1.36

war

1.33

OPS

1.28

Wars

1.18

ERA

1.17

 Wars

1.14

 FANTASY

1.11

War

1.11

Activations Density 0.013%

keywords related to wars and conflicts

references to war

No Comments

No Known Activations