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

urga

-0.78

ctuary

-0.75

ilon

-0.73

ource

-0.68

å£«

-0.67

ylum

-0.66

 Outbreak

-0.66

ivil

-0.65

Mobil

-0.64

aepernick

-0.63

POSITIVE LOGITS

 âĶľ

0.83

 framing

0.63

 Dove

0.61

_>

0.61

plets

0.59

EST

0.59

dq

0.57

 suspended

0.57

 Dustin

0.57

uable

0.57

Activations Density 0.000%

No Known Activations

This feature has no known activations.