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

ullah

-0.72

utterstock

-0.71

reatment

-0.71

 illustration

-0.67

 Imam

-0.63

 neighbor

-0.62

aiman

-0.62

gregation

-0.61

lyn

-0.61

 biking

-0.61

POSITIVE LOGITS

 Shinra

0.73

Leaks

0.73

 Explosive

0.71

ãĥ´ãĤ¡

0.71

CV

0.69

 Enemies

0.69

igne

0.69

 myster

0.69

addons

0.68

ãĢĳ

0.68

Activations Density 0.000%

No Known Activations

This feature has no known activations.