INDEX

Explanations

mentions of torture in various contexts

oai_token-act-pair · gpt-3.5-turbo

references to torture and related abuses

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

soType

-0.80

ership

-0.75

ijk

-0.65

soDeliveryDate

-0.65

arger

-0.65

ovember

-0.63

nect

-0.63

 explan

-0.63

utsch

-0.63

Merit

-0.62

POSITIVE LOGITS

 torture

0.98

 tortured

0.77

 captives

0.75

 tactics

0.73

imony

0.73

 detainees

0.72

 confinement

0.72

apons

0.70

 torment

0.70

rs

0.69

Activations Density 0.021%

mentions of torture in various contexts

references to torture and related abuses

No Comments

No Known Activations