INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

veyard

-0.77

hetto

-0.75

jug

-0.70

gage

-0.69

ibling

-0.69

lict

-0.66

ffiti

-0.66

uction

-0.65

 Merchants

-0.63

isine

-0.63

POSITIVE LOGITS

>>>

0.78

rake

0.72

sie

0.70

pas

0.68

antine

0.67

SPONSORED

0.66

_>

0.64

bis

0.64

embed

0.63

ruby

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.