INDEX

Explanations

specific references to "AS" acronym

oai_token-act-pair · gpt-3.5-turbo

references to specific organizations or entities, particularly focusing on the acronym "AS."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 Britann

-0.72

 Schne

-0.68

 MacArthur

-0.66

naires

-0.63

town

-0.62

endor

-0.60

>>\

-0.60

 Buffett

-0.59

ogue

-0.59

 Destroyer

-0.59

POSITIVE LOGITS

ADA

1.05

ync

1.03

ynchronous

1.01

leep

1.00

daq

0.99

sembly

0.96

AS

0.95

AN

0.94

ENA

0.94

ANA

0.94

Activations Density 0.008%

specific references to "AS" acronym

references to specific organizations or entities, particularly focusing on the acronym "AS."

No Comments

No Known Activations