INDEX

Explanations

numeric ranges or lists

oai_token-act-pair · gpt-3.5-turbo

references to official procedures or registrations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

èĢħ

-0.49

 Mehran

-0.46

Äĩ

-0.44

 hesitated

-0.44

 reverted

-0.43

 lasted

-0.41

 kidding

-0.41

 notwithstanding

-0.40

vou

-0.40

ãĥ¼ãĥĨãĤ£

-0.40

POSITIVE LOGITS

teness

0.42

doms

0.42

ships

0.41

ties

0.39

 vantage

0.39

power

0.39

mph

0.39

ulz

0.39

arent

0.38

ds

0.37

Activations Density 2.845%

numeric ranges or lists

references to official procedures or registrations

No Comments

No Known Activations