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

 municip

-0.72

 sanitation

-0.66

 fire

-0.61

ÃĥÃĤ

-0.61

 Purg

-0.60

san

-0.59

 surroundings

-0.59

mM

-0.58

vot

-0.58

 yourselves

-0.57

POSITIVE LOGITS

emort

0.84

emic

0.77

ulhu

0.76

enei

0.73

heid

0.73

peed

0.73

 Eternity

0.71

ylum

0.70

ftime

0.69

exclusive

0.69

Activations Density 0.000%

No Known Activations

This feature has no known activations.