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

rica

-0.91

Medium

-0.88

ovi

-0.87

apest

-0.83

yrinth

-0.83

anwhile

-0.81

apo

-0.72

ibu

-0.70

ln

-0.67

nexus

-0.67

POSITIVE LOGITS

jit

0.62

inker

0.60

hered

0.59

uous

0.58

Wen

0.58

ille

0.58

yll

0.58

icht

0.57

Er

0.57

ple

0.56

Activations Density 0.000%

No Known Activations

This feature has no known activations.