INDEX

Explanations

verbs or nouns related to causing harm or damage

oai_token-act-pair · gpt-3.5-turbo

words related to destruction or damage

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

DragonMagazine

-0.87

uchs

-0.74

zbek

-0.70

gomery

-0.69

ussen

-0.68

uana

-0.67

acus

-0.66

liner

-0.66

KER

-0.65

glas

-0.64

POSITIVE LOGITS

 havoc

1.34

roying

0.83

adoes

0.82

 wrought

0.82

 arte

0.73

 habitats

0.73

wre

0.73

 nests

0.71

 piles

0.71

ados

0.71

Activations Density 0.092%

verbs or nouns related to causing harm or damage

words related to destruction or damage

No Comments

No Known Activations