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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

perty

-0.83

quo

-0.67

estial

-0.67

ependent

-0.64

auri

-0.64

 Realms

-0.64

 Patient

-0.63

monds

-0.63

perial

-0.63

raft

-0.63

POSITIVE LOGITS

 Shiny

0.72

ãĤ¼ãĤ¦ãĤ¹

0.69

æ©

0.68

 congr

0.67

DX

0.67

ãĤ·

0.67

CPC

0.65

å°Ĩ

0.65

inc

0.65

ousse

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.