INDEX

Explanations

proper nouns related to locations or individuals

oai_token-act-pair · gpt-3.5-turbo

references to specific individuals or entities, particularly names that start with "Cort."

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

lihood

-0.73

Gi

-0.68

 conceivable

-0.66

 paranormal

-0.65

 Freak

-0.65

 germ

-0.65

 bang

-0.64

 fruitful

-0.63

Magikarp

-0.63

ÏĦ

-0.63

POSITIVE LOGITS

icol

1.14

icum

1.00

Ã©s

0.98

heast

0.96

ez

0.91

rance

0.87

 cort

0.87

oise

0.87

uning

0.85

isol

0.85

Activations Density 0.019%

proper nouns related to locations or individuals

references to specific individuals or entities, particularly names that start with "Cort."

No Comments

No Known Activations