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

 Attribution

-0.74

 Lavrov

-0.66

aste

-0.65

colour

-0.62

voy

-0.61

 sensit

-0.60

 programme

-0.60

 Exper

-0.60

Fav

-0.59

ande

-0.59

POSITIVE LOGITS

eeks

0.73

Magikarp

0.72

rences

0.72

ħĭ

0.68

ayson

0.65

annis

0.64

DoS

0.64

 Klux

0.63

EStream

0.62

ulhu

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.