INDEX

Explanations

positive descriptions about someone's character and abilities

oai_token-act-pair · gpt-3.5-turbo

attributes describing exceptional qualities in 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

 Worse

-0.73

levard

-0.70

Bye

-0.67

rogens

-0.65

 Spectre

-0.64

Ń·

-0.63

 Secondary

-0.62

 Parenthood

-0.62

 balk

-0.62

 Sabha

-0.61

POSITIVE LOGITS

 intellect

1.05

 charisma

1.03

 knack

0.97

 demeanor

0.96

 personality

0.92

ahime

0.91

 witty

0.89

 compassion

0.89

 unmatched

0.87

 temperament

0.86

Activations Density 0.381%

positive descriptions about someone's character and abilities

attributes describing exceptional qualities in individuals

No Comments

No Known Activations