INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Oaks

-0.89

Gas

-0.74

 bowls

-0.68

 dispatcher

-0.67

 baskets

-0.63

 Sandy

-0.62

 eclipse

-0.62

athi

-0.61

 observer

-0.59

 Anita

-0.59

POSITIVE LOGITS

GY

0.76

today

0.71

 ..........

0.67

Goal

0.66

fighters

0.66

=~

0.66

chance

0.65

reme

0.65

 ðŁ

0.64

Ã§a

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.