INDEX

Explanations

phrases related to the name "Benjamin" which is a clue to its behavior, probably focusing on specific texts that mention the name "Benjamin"

oai_token-act-pair · gpt-3.5-turbo

mentions of the name "Benjamin" and variations of it

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

ebook

-0.90

Reviewer

-0.88

uid

-0.83

atics

-0.82

atically

-0.80

monary

-0.76

ticket

-0.75

TOR

-0.74

govtrack

-0.74

atic

-0.73

POSITIVE LOGITS

jamin

1.39

 Netanyahu

1.17

 Franklin

1.15

 Benjamin

0.93

 Centauri

0.86

 Witt

0.83

itude

0.79

 Tucker

0.76

heimer

0.75

 Harrison

0.74

Activations Density 0.005%

phrases related to the name "Benjamin" which is a clue to its behavior, probably focusing on specific texts that mention the name "Benjamin"

mentions of the name "Benjamin" and variations of it

No Comments

No Known Activations