INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

 stripes

-0.71

Initialized

-0.69

 floral

-0.65

ItemTracker

-0.63

IL

-0.61

Ø§Ø

-0.61

bra

-0.60

rency

-0.59

Closure

-0.58

 Richardson

-0.58

POSITIVE LOGITS

adle

0.74

intend

0.72

ollah

0.69

atches

0.69

rahim

0.69

 econom

0.68

ulously

0.68

oru

0.66

 Nanto

0.65

zin

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.