INDEX

Explanations

phrases related to receiving awards or rewards

oai_token-act-pair · gpt-3.5-turbo

instances of the word "awarded" and its variations, indicating recognition or prizes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

ierre

-0.78

loo

-0.73

bian

-0.67

andem

-0.66

ben

-0.65

 Mong

-0.62

ijn

-0.61

xon

-0.61

ydia

-0.60

cp

-0.60

POSITIVE LOGITS

 prizes

1.07

 awarded

1.03

 awards

0.98

 awarding

0.95

Winner

0.94

 award

0.83

 laure

0.79

 prize

0.78

Reviewer

0.76

 recipients

0.74

Activations Density 0.022%

phrases related to receiving awards or rewards

instances of the word "awarded" and its variations, indicating recognition or prizes

No Comments

No Known Activations