INDEX

Explanations

instructions or directives related to enabling or using certain features on a website

oai_token-act-pair · gpt-3.5-turbo

calls to action related to enabling features or providing support

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

MRI

-0.80

pan

-0.76

ailability

-0.72

cular

-0.70

capacity

-0.70

angled

-0.69

rium

-0.67

ilogy

-0.67

olver

-0.66

azeera

-0.66

POSITIVE LOGITS

 thous

0.81

sir

0.81

 Submit

0.78

 discretion

0.76

ername

0.74

 thank

0.72

 acknowled

0.72

 PLEASE

0.70

 subscrib

0.69

 sacrific

0.68

Activations Density 0.210%

instructions or directives related to enabling or using certain features on a website

calls to action related to enabling features or providing support

No Comments

No Known Activations