INDEX

Explanations

proper nouns related to politics or business

oai_token-act-pair · gpt-3.5-turbo

references to specific individuals and industries

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

 Blaz

-0.75

ãĥīãĥ©ãĤ´ãĥ³

-0.73

 TAMADRA

-0.72

é»Ĵ

-0.67

 Shed

-0.65

OGR

-0.64

DOC

-0.63

 mosqu

-0.63

NX

-0.62

olved

-0.62

POSITIVE LOGITS

rial

1.17

rian

1.09

rine

1.06

rator

1.05

rum

1.03

rations

1.02

rative

1.01

ria

1.00

ri

0.99

ris

0.99

Activations Density 0.043%

proper nouns related to politics or business

references to specific individuals and industries

No Comments

No Known Activations