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

-0.84

ãĥĥ

-0.72

 derogatory

-0.71

ËĪ

-0.68

 conve

-0.62

ãĥ³

-0.62

 quant

-0.60

xual

-0.60

 buds

-0.59

 dates

-0.59

POSITIVE LOGITS

illac

0.75

avia

0.71

osate

0.70

htaking

0.69

raltar

0.68

anski

0.68

gerald

0.67

STL

0.67

fman

0.67

SON

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.