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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

emonium

-0.90

 kits

-0.74

 uniformly

-0.73

nesota

-0.68

 scoring

-0.68

 sorely

-0.67

 peril

-0.66

 leveling

-0.64

 locking

-0.63

ful

-0.62

POSITIVE LOGITS

ieu

0.83

ando

0.78

ivo

0.77

brush

0.77

armac

0.76

_(

0.72

Pic

0.70

 Verse

0.70

adr

0.69

Direct

0.69

Activations Density 0.000%

No Known Activations

This feature has no known activations.