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

reon

-0.75

ichael

-0.69

TeX

-0.68

 prov

-0.64

lette

-0.61

ulet

-0.60

Staff

-0.59

 Â¯

-0.59

anguage

-0.59

 stake

-0.58

POSITIVE LOGITS

é¾įå

0.85

ontent

0.73

eatures

0.71

eches

0.65

tested

0.65

 Ambro

0.64

Osw

0.64

ivating

0.64

icably

0.63

Īè

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.