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

 Nieto

-0.73

 Vide

-0.72

 Rend

-0.70

ness

-0.68

ional

-0.68

 VIDEOS

-0.67

MEN

-0.66

 Dragonbound

-0.65

 Priv

-0.64

Que

-0.63

POSITIVE LOGITS

rosis

0.89

acid

0.84

icter

0.81

obook

0.80

ffiti

0.75

Sov

0.73

ritis

0.73

irie

0.72

ashtra

0.72

onite

0.71

Activations Density 0.000%

No Known Activations

This feature has no known activations.