INDEX

Explanations

phrases related to receiving awards or prizes

oai_token-act-pair · gpt-3.5-turbo

instances of the word "awarded."

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

ierre

-0.78

ben

-0.67

EEK

-0.66

-0.64

bian

-0.64

APD

-0.63

ydia

-0.63

so

-0.63

loo

-0.63

sen

-0.63

POSITIVE LOGITS

 awarded

1.01

 prizes

1.00

 awards

0.97

 awarding

0.91

 award

0.83

Winner

0.80

 accol

0.79

 laure

0.78

Reviewer

0.72

 prize

0.71

Activations Density 0.019%

phrases related to receiving awards or prizes

instances of the word "awarded."

No Comments

No Known Activations