INDEX

Explanations

references to armed forces, military activities, and related terms

oai_token-act-pair · gpt-3.5-turbo

instances of the word "armed."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

rx

-0.96

article

-0.78

MAL

-0.78

rb

-0.78

RGB

-0.72

HUD

-0.71

 Remastered

-0.71

arget

-0.70

 Woodward

-0.69

MRI

-0.68

POSITIVE LOGITS

 robbery

0.96

 guards

0.95

 armed

0.94

 insurrection

0.92

 uprising

0.84

ament

0.84

 robbers

0.82

 robber

0.80

 robberies

0.80

 revolt

0.80

Activations Density 0.021%

references to armed forces, military activities, and related terms

instances of the word "armed."

No Comments

No Known Activations