INDEX

Explanations

pronouns and verbs related to informing or alerting others

oai_token-act-pair · gpt-3.5-turbo

phrases and words related to communication or notifications to individuals

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

agne

-0.69

evil

-0.59

laughter

-0.59

 analogy

-0.59

BuyableInstoreAndOnline

-0.58

Laughs

-0.56

 Purg

-0.55

nan

-0.54

 Murd

-0.54

è¦

-0.54

POSITIVE LOGITS

 beforehand

1.10

 about

1.03

 whereabouts

0.96

 thereof

0.91

 orally

0.86

about

0.86

 regarding

0.83

via

0.83

 ABOUT

0.81

 promptly

0.79

Activations Density 0.162%

pronouns and verbs related to informing or alerting others

phrases and words related to communication or notifications to individuals

No Comments

No Known Activations