INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

code

-0.74

İ

-0.68

owder

-0.67

ounce

-0.66

opoly

-0.65

Mods

-0.61

 mischief

-0.61

Lay

-0.60

 calendars

-0.59

 ILCS

-0.59

POSITIVE LOGITS

%"

0.85

ãĢı

0.76

Sov

0.76

olkien

0.74

ãģĤ

0.72

ãĤĮ

0.72

alore

0.68

itbart

0.65

|--

0.65

?'"

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.