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

 appl

-0.71

enz

-0.70

æĪ¦

-0.68

atom

-0.67

aceous

-0.63

 Grimoire

-0.63

rio

-0.63

 Meteor

-0.60

culosis

-0.59

 Aven

-0.59

POSITIVE LOGITS

writing

0.76

orsi

0.74

mbuds

0.74

abouts

0.72

gres

0.67

fore

0.66

write

0.65

marriage

0.65

gap

0.64

eve

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.