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

ensibly

-0.77

olean

-0.74

aquin

-0.70

ãģ¦

-0.70

iosyncr

-0.69

 Mehran

-0.68

 destro

-0.67

bably

-0.67

thouse

-0.66

mble

-0.65

POSITIVE LOGITS

Own

0.72

lear

0.65

holder

0.65

 Nurs

0.63

 Fitzgerald

0.63

hold

0.62

nia

0.62

Dad

0.62

ayn

0.61

 Loch

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.