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

å¥³

-0.71

 CLIENT

-0.67

Interstitial

-0.67

Scar

-0.66

RC

-0.65

 unseen

-0.65

Ake

-0.65

KA

-0.64

 Alleg

-0.59

 Aurora

-0.59

POSITIVE LOGITS

alus

0.85

alist

0.83

ratulations

0.77

amination

0.75

iversal

0.70

laus

0.69

afa

0.69

iquid

0.67

hyde

0.67

pend

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.