INDEX

Explanations

expressions related to giving or presenting something to someone

oai_token-act-pair · gpt-3.5-turbo

phrases that include the word "a" followed by various descriptors

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

ources

-0.75

ami

-0.72

Austral

-0.70

effects

-0.69

aneously

-0.67

AI

-0.66

Ct

-0.65

 Effect

-0.65

 trends

-0.65

 suicides

-0.65

POSITIVE LOGITS

 chance

1.21

 choice

0.94

 gift

0.92

 rundown

0.91

 headache

0.91

 berth

0.88

 clue

0.87

 breat

0.86

 haircut

0.86

 heads

0.86

Activations Density 0.127%

expressions related to giving or presenting something to someone

phrases that include the word "a" followed by various descriptors

No Comments

No Known Activations