INDEX

Explanations

mentions of the name "Jordan."

oai_token-act-pair · gpt-3.5-turbo

the name "Jordan."

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

ntil

-0.77

forms

-0.73

emetery

-0.72

ancial

-0.71

ULAR

-0.71

 nutrit

-0.71

 redund

-0.68

roup

-0.68

teenth

-0.68

ulum

-0.67

POSITIVE LOGITS

 Jordan

1.10

Jordan

0.95

 Peterson

0.92

olson

0.84

sey

0.81

 Henderson

0.80

oglu

0.80

allah

0.77

pora

0.77

endish

0.75

Activations Density 0.012%

mentions of the name "Jordan."

the name "Jordan."

No Comments

No Known Activations