INDEX

Explanations

phrases related to critical evaluations or assessments within a context

oai_token-act-pair · gpt-3.5-turbo

ellipses indicating pauses or breaks in thought

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

–

-0.94

.–

-0.81

Advertisements

-0.80

————————————————

-0.65

âĢĲ

-0.65

agame

-0.62

—-

-0.60

)—

-0.58

*/(

-0.57

theless

-0.54

POSITIVE LOGITS

...

3.64

 ..."

2.99

 ....

2.95

 ......

2.66

 [...]

2.47

,...

2.30

..

2.27

...

2.24

...]

2.23

 âĢ¦

2.22

Activations Density 0.012%

phrases related to critical evaluations or assessments within a context

ellipses indicating pauses or breaks in thought

No Comments

No Known Activations