INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

inea

-0.75

Dug

-0.69

 wand

-0.68

rov

-0.64

uces

-0.64

 streaks

-0.62

ivated

-0.62

abilia

-0.62

 Mayhem

-0.62

ificent

-0.62

POSITIVE LOGITS

ãĥĨãĤ£

0.78

 ende

0.71

pipe

0.69

ĸļ

0.67

Catal

0.64

hani

0.64

grave

0.64

FORE

0.63

meier

0.63

anguages

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.