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

QUI

-0.80

OLOG

-0.71

Topic

-0.70

tag

-0.68

qq

-0.65

OPLE

-0.65

TAG

-0.65

OLOGY

-0.64

sie

-0.63

ļéĨĴ

-0.63

POSITIVE LOGITS

nces

0.79

yre

0.75

chwitz

0.72

hement

0.70

 Painter

0.67

 Lauder

0.66

burning

0.65

skirts

0.63

iasm

0.62

dil

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.