INDEX

Explanations

phrases related to the navigation app Waze

oai_token-act-pair · gpt-3.5-turbo

words related to enjoyment or positive experiences

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

sburgh

-0.78

risome

-0.77

ITIES

-0.75

erest

-0.72

arians

-0.67

ulates

-0.67

rim

-0.65

ulators

-0.65

riage

-0.65

ancial

-0.61

POSITIVE LOGITS

 Flavoring

0.88

ecake

0.81

aze

0.79

ze

0.76

eker

0.76

zes

0.74

ey

0.73

;;;;;;;;;;;;

0.72

eb

0.71

quit

0.70

Activations Density 0.014%

phrases related to the navigation app Waze

words related to enjoyment or positive experiences

No Comments

No Known Activations