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

cyclopedia

-0.77

ilib

-0.73

oding

-0.72

ween

-0.70

apse

-0.69

 Jagu

-0.68

bda

-0.66

LT

-0.66

iang

-0.66

Ct

-0.66

POSITIVE LOGITS

pas

0.66

pheus

0.65

backs

0.64

 insanity

0.64

raped

0.63

 Sessions

0.61

 shove

0.61

aukee

0.60

Americans

0.60

Reloaded

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.