INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

kefeller

-0.78

 Volunte

-0.78

enegger

-0.76

ieved

-0.73

 Reborn

-0.72

 Nicarag

-0.71

 sidx

-0.69

untled

-0.68

ipient

-0.67

geoning

-0.67

POSITIVE LOGITS

 bulk

0.70

 quote

0.68

 hairc

0.64

ear

0.63

 over

0.62

 halftime

0.61

 denominations

0.61

 Price

0.61

overs

0.60

 caps

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.