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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ilege

-0.77

abilities

-0.73

Mi

-0.69

 tant

-0.62

 Item

-0.62

insula

-0.62

sag

-0.61

controller

-0.60

metadata

-0.59

 disclosure

-0.59

POSITIVE LOGITS

Ö¼

0.94

ally

0.84

obyl

0.78

oslov

0.74

hift

0.74

 diplom

0.73

×Ļ×

0.72

OHN

0.72

tein

0.72

ynt

0.71

Activations Density 0.000%

No Known Activations

This feature has no known activations.