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

 Testament

-0.68

%]

-0.65

cellence

-0.64

OGR

-0.60

escription

-0.60

_-

-0.60

icism

-0.60

 orbits

-0.60

-0.59

hemy

-0.59

POSITIVE LOGITS

 household

1.15

enter

0.71

DAQ

0.67

olute

0.66

shore

0.64

phal

0.64

Accessory

0.63

auc

0.62

 homeowners

0.61

ha

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.