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

Alert

-0.79

earchers

-0.73

formed

-0.71

iliated

-0.67

forming

-0.66

fters

-0.64

ateful

-0.63

lishing

-0.63

ahead

-0.63

ceptive

-0.62

POSITIVE LOGITS

itaire

0.82

ãĤ©

0.74

OTA

0.73

izon

0.72

rolet

0.70

ilan

0.69

 weeds

0.69

ï¿½ï¿½ï¿½ï¿½

0.68

aroo

0.67

inav

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.