INDEX

Explanations

calls to action for engaging with newsletters or subscriptions

oai_token-act-pair · gpt-3.5-turbo

imperative verbs related to actions or calls to action

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

][

-0.66

ortium

-0.65

ippi

-0.65

eni

-0.63

swick

-0.62

ensation

-0.59

uum

-0.58

olit

-0.58

edge

-0.58

roma

-0.58

POSITIVE LOGITS

 yourself

1.10

 yourselves

1.04

 accordingly

0.92

 alike

0.82

 whichever

0.76

 ourselves

0.72

 oneself

0.72

 Yourself

0.71

 your

0.70

 versa

0.66

Activations Density 0.154%

calls to action for engaging with newsletters or subscriptions

imperative verbs related to actions or calls to action

No Comments

No Known Activations