INDEX

Explanations

words that denote importance, specificity, or relevance related to different contexts or subjects

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

asus

-0.63

akedown

-0.61

olicy

-0.56

olutions

-0.56

onut

-0.54

inances

-0.54

UID

-0.52

 Pump

-0.51

 Mant

-0.50

hower

-0.50

POSITIVE LOGITS

 enough

1.17

 unto

1.08

 thereto

1.06

izable

1.05

to

1.03

ially

1.02

enough

0.99

isable

0.99

ibly

0.98

ãĥ¼ãĥĨ

0.97

Activations Density 0.288%

words that denote importance, specificity, or relevance related to different contexts or subjects

No Comments

No Known Activations