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

abba

-0.72

eals

-0.71

hus

-0.69

iseum

-0.68

igate

-0.68

raid

-0.67

erry

-0.66

undle

-0.66

ivo

-0.66

naissance

-0.65

POSITIVE LOGITS

Nit

0.69

 Kyoto

0.63

 Jacket

0.61

Reply

0.57

ãĤ¶

0.57

 chron

0.57

Ö¼

0.57

 Influ

0.56

REDACTED

0.56

Applications

0.56

Activations Density 0.000%

No Known Activations

This feature has no known activations.