INDEX

Explanations

words related to negativity or negation

oai_token-act-pair · gpt-3.5-turbo

terms associated with negativity or neglect

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

realDonaldTrump

-0.75

BuyableInstoreAndOnline

-0.67

 Falk

-0.65

WATCH

-0.65

 Mour

-0.65

VEL

-0.63

 Takeru

-0.62

 Rath

-0.62

ORGE

-0.61

 Butterfly

-0.61

POSITIVE LOGITS

otiation

1.71

oti

1.68

atives

1.40

ativity

1.24

rito

1.20

lected

1.20

lig

1.15

atively

1.14

ative

1.11

oci

1.08

Activations Density 0.020%

words related to negativity or negation

terms associated with negativity or neglect

No Comments

No Known Activations