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

trak

-0.90

git

-0.80

liga

-0.76

hematic

-0.73

agos

-0.69

zik

-0.68

phalt

-0.68

xious

-0.66

TPPStreamerBot

-0.66

stem

-0.66

POSITIVE LOGITS

 Vaughn

0.68

enegger

0.67

vo

0.65

Lun

0.64

 Fellowship

0.64

 Practices

0.63

lyn

0.60

Begin

0.60

 repr

0.60

 Thieves

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.