INDEX

Explanations

references to a specific location, likely "Minneapolis."

oai_token-act-pair · gpt-3.5-turbo

references to the word "Min" and its variations, indicating a focus on a specific entity or theme associated with that term

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

enegger

-0.85

 Compass

-0.64

red

-0.64

atered

-0.60

 Cascade

-0.60

 Crusade

-0.60

 centre

-0.60

 almonds

-0.60

 revolving

-0.60

 links

-0.59

POSITIVE LOGITS

iatures

1.18

imal

1.14

neapolis

1.07

erva

1.02

nesota

1.01

oan

1.01

imize

1.00

utes

0.99

amoto

0.98

ority

0.97

Activations Density 0.013%

references to a specific location, likely "Minneapolis."

references to the word "Min" and its variations, indicating a focus on a specific entity or theme associated with that term

No Comments

No Known Activations