INDEX

Explanations

phrases related to destruction and conflict

oai_token-act-pair · gpt-3.5-turbo

instances of destruction or harm in contexts related to conflict

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

ITNESS

-0.69

PLIED

-0.68

 Assistance

-0.64

 Representative

-0.63

ç·

-0.63

irin

-0.63

 Lithuan

-0.62

 Dating

-0.62

ogle

-0.61

>[

-0.61

POSITIVE LOGITS

 favor

0.84

 senseless

0.81

pload

0.75

 merciless

0.73

 favour

0.72

 scourge

0.71

 sheer

0.71

 brutally

0.70

 entirety

0.69

 manageable

0.68

Activations Density 0.483%

phrases related to destruction and conflict

instances of destruction or harm in contexts related to conflict

No Comments

No Known Activations