INDEX

Explanations

phrases related to actions and activities

oai_token-act-pair · gpt-3.5-turbo

connection words and phrases related to grouping and organization

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

ãĤ¹ãĥĪ

-0.71

iolet

-0.68

bern

-0.67

cht

-0.67

nex

-0.66

mx

-0.66

tec

-0.66

ohl

-0.66

atom

-0.66

cit

-0.66

POSITIVE LOGITS

 placing

1.38

 treating

1.34

 transferring

1.31

 applying

1.31

 putting

1.31

 releasing

1.30

 settling

1.30

 taking

1.30

 registering

1.29

 stopping

1.29

Activations Density 0.164%

phrases related to actions and activities

connection words and phrases related to grouping and organization

No Comments

No Known Activations