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

jong

-0.68

ional

-0.65

inson

-0.64

OTUS

-0.64

 guiActiveUnfocused

-0.61

Timeout

-0.60

 escalate

-0.60

hare

-0.59

eday

-0.59

 Jinping

-0.59

POSITIVE LOGITS

anchester

0.68

ARGET

0.65

irtual

0.65

isers

0.65

GoldMagikarp

0.65

Eva

0.64

 Savannah

0.64

estate

0.63

kinson

0.63

 Glasgow

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.