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

osate

-0.80

isine

-0.79

yip

-0.74

otine

-0.71

 disapp

-0.71

buster

-0.69

 Meal

-0.69

odon

-0.67

oche

-0.66

tick

-0.66

POSITIVE LOGITS

emin

0.73

 backups

0.73

 corners

0.70

 volumes

0.65

 trusts

0.64

 delicate

0.63

 colleges

0.63

 switches

0.63

enture

0.63

 sclerosis

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.