INDEX

Explanations

the word "empty."

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "empty."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

TM

-0.74

xual

-0.72

-0.69

Â®

-0.65

conom

-0.65

ivals

-0.64

 Factor

-0.63

Â®,

-0.63

beam

-0.62

 relates

-0.62

POSITIVE LOGITS

 empty

3.66

empty

2.82

 Empty

2.54

Empty

2.28

 emptied

1.99

 vacant

1.94

 empt

1.91

 emptiness

1.89

 unused

1.63

 blank

1.53

Activations Density 0.016%

the word "empty."

occurrences of the word "empty."

No Comments

No Known Activations