INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

igraph

-0.82

osi

-0.77

ski

-0.72

 Tyson

-0.68

ukong

-0.68

osaurs

-0.67

 mornings

-0.66

stories

-0.66

 Grimm

-0.65

 sights

-0.65

POSITIVE LOGITS

terness

0.68

ologne

0.65

pex

0.64

oline

0.64

olate

0.62

uxe

0.60

âĺħâĺħ

0.59

VW

0.58

LIB

0.58

 Facility

0.57

Activations Density 0.000%

No Known Activations

This feature has no known activations.