INDEX

Explanations

phrases emphasizing certainty or affirmation

oai_token-act-pair · gpt-3.5-turbo

reaffirmations of certainty

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

 Rolls

-0.80

mare

-0.75

ĸļ

-0.73

 Dull

-0.66

orate

-0.63

 Span

-0.60

 Duchess

-0.59

pat

-0.59

 Notting

-0.57

 Riders

-0.57

POSITIVE LOGITS

 sake

1.28

 purposes

1.04

ummies

0.94

 reasons

0.92

iatus

0.68

farious

0.68

 Reasons

0.66

ansom

0.63

 overdue

0.63

 undisclosed

0.60

Activations Density 0.118%

phrases emphasizing certainty or affirmation

reaffirmations of certainty

No Comments

No Known Activations