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

 Rockefeller

-0.72

 Freem

-0.69

 cult

-0.67

ho

-0.66

 Prometheus

-0.65

 Kickstarter

-0.64

 Philos

-0.63

 Roma

-0.62

Tao

-0.62

 Gong

-0.61

POSITIVE LOGITS

ĸļ

0.76

zens

0.72

pora

0.70

ombat

0.70

ãĤ¼ãĤ¦ãĤ¹

0.69

soever

0.67

obar

0.66

ameda

0.65

ascript

0.64

oute

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.