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

rait

-0.86

 Leban

-0.80

 Malf

-0.79

Elk

-0.74

lyak

-0.74

 Dull

-0.68

Eg

-0.65

McA

-0.65

mage

-0.63

Fri

-0.63

POSITIVE LOGITS

¿½

0.69

ģĸ

0.67

 surprises

0.66

Ĥª

0.66

rophic

0.65

 needles

0.64

rates

0.63

 recipients

0.62

atin

0.62

é¾į

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.