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

ishable

-0.95

 partic

-0.70

forming

-0.68

 inexpensive

-0.67

 nascent

-0.67

 narc

-0.65

 specialized

-0.65

ger

-0.65

icipated

-0.64

obin

-0.63

POSITIVE LOGITS

 Skydragon

0.83

FontSize

0.79

folk

0.76

 Mania

0.75

ARK

0.73

osate

0.72

\\\\\\\\

0.71

################

0.70

ãĤ¦ãĤ¹

0.69

 ILCS

0.69

Activations Density 0.000%

No Known Activations

This feature has no known activations.