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

Assistant

-0.73

>>>>>>>>

-0.71

Ħ¢

-0.68

 extrad

-0.66

iT

-0.64

 autistic

-0.63

assing

-0.63

MRI

-0.63

hyde

-0.62

thia

-0.60

POSITIVE LOGITS

chieve

0.80

 Quadro

0.69

oes

0.64

quished

0.63

daq

0.63

ril

0.62

chev

0.61

ournals

0.61

 Jarrett

0.60

andals

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.