INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 cannons

-0.70

iaries

-0.69

 dracon

-0.68

housing

-0.67

 Herald

-0.67

 cannon

-0.66

aer

-0.64

ITS

-0.59

techn

-0.59

isitions

-0.59

POSITIVE LOGITS

 âĶľ

0.81

ando

0.74

arov

0.72

 Bulg

0.72

uel

0.71

Ð²

0.71

hov

0.71

 âĶľâĶĢâĶĢ

0.67

coe

0.66

henko

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.