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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

catentry

-0.75

Thumbnail

-0.73

iband

-0.72

Emer

-0.67

avascript

-0.67

ãĤ¼ãĤ¦ãĤ¹

-0.67

CRIP

-0.66

bye

-0.65

ãĥ¼ãĥĨãĤ£

-0.64

 Integration

-0.64

POSITIVE LOGITS

seless

0.73

thing

0.70

wise

0.66

ological

0.66

 Bender

0.65

 titles

0.64

 convictions

0.63

rage

0.62

QUEST

0.61

rants

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.