INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

00200000

-0.83

ONSORED

-0.82

vati

-0.75

ham

-0.66

ampa

-0.66

AMS

-0.66

agine

-0.63

uates

-0.63

)=(

-0.62

escription

-0.61

POSITIVE LOGITS

 Argent

0.65

 Quan

0.63

oret

0.63

Versions

0.62

Wid

0.61

 Huss

0.61

roma

0.59

otropic

0.59

pring

0.58

 neigh

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.