INDEX

Explanations

phrases indicating possession or attribution of an action or characteristic

oai_token-act-pair · gpt-3.5-turbo

instances of the word "had."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

bie

-0.64

hammer

-0.59

ocol

-0.59

ensing

-0.58

ieg

-0.57

porting

-0.57

coin

-0.57

umping

-0.57

 Voters

-0.56

 Skies

-0.56

POSITIVE LOGITS

 been

1.26

 undergone

1.15

 begun

1.02

 gone

1.02

iths

1.00

 gotten

0.94

 taken

0.91

 become

0.90

raltar

0.87

 come

0.85

Activations Density 0.158%

phrases indicating possession or attribution of an action or characteristic

instances of the word "had."

No Comments

No Known Activations