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

gur

-0.76

 Guer

-0.74

MET

-0.73

è£

-0.71

¬

-0.69

Face

-0.65

FACE

-0.65

Pixel

-0.65

ensor

-0.63

Ø¨

-0.61

POSITIVE LOGITS

cffff

0.82

valued

0.74

kind

0.74

theless

0.72

 challeng

0.72

 incorpor

0.71

taboola

0.71

 pestic

0.68

eless

0.68

otle

0.68

Activations Density 0.000%

No Known Activations

This feature has no known activations.