INDEX

Explanations

instances where there is an exception or deviation from the norm

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate exceptions or deviations from the norm

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

 horizont

-0.66

nesday

-0.65

 Majesty

-0.63

Keys

-0.62

 Coffee

-0.60

foot

-0.59

 scrimmage

-0.58

anon

-0.58

 canvas

-0.57

Sta

-0.57

POSITIVE LOGITS

 behaved

0.80

worldly

0.75

heastern

0.73

 inclined

0.69

omon

0.68

é¾įåĸļå£«

0.67

NESS

0.66

ghazi

0.65

scill

0.64

OSED

0.63

Activations Density 0.019%

instances where there is an exception or deviation from the norm

phrases that indicate exceptions or deviations from the norm

No Comments

No Known Activations