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

drop

-0.74

lake

-0.68

gall

-0.66

QC

-0.65

ben

-0.63

llor

-0.62

qt

-0.62

Italy

-0.61

 Angus

-0.61

 laureate

-0.59

POSITIVE LOGITS

 sidx

0.77

 Immunity

0.69

oneliness

0.67

itely

0.67

yrics

0.66

 specificity

0.65

iants

0.65

abbit

0.65

.):

0.65

 Spells

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.