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

asaki

-0.69

åº

-0.69

 resid

-0.67

eca

-0.66

 delegates

-0.66

usra

-0.65

 Claudia

-0.64

eria

-0.63

CPC

-0.63

 sails

-0.60

POSITIVE LOGITS

gebra

0.64

âĸĳ

0.62

 Types

0.61

gue

0.61

 Reich

0.60

gart

0.59

 Wonders

0.58

swers

0.58

 Intelligence

0.58

 Dogs

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.