INDEX

Explanations

phrases related to armed forces and military activities

oai_token-act-pair · gpt-3.5-turbo

mentions of armed forces or armed individuals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

-0.82

MAL

-0.76

rb

-0.74

Bee

-0.72

Publisher

-0.71

HUD

-0.70

VEN

-0.69

article

-0.69

TX

-0.69

Â¢

-0.68

POSITIVE LOGITS

 armed

1.02

 robbery

0.96

 uprising

0.93

 guards

0.91

 insurrection

0.88

 revolt

0.87

 robber

0.86

 arming

0.85

 robbers

0.84

 conflic

0.82

Activations Density 0.012%

phrases related to armed forces and military activities

mentions of armed forces or armed individuals

No Comments

No Known Activations