INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

pring

-0.87

weeney

-0.70

aturday

-0.68

 moderation

-0.66

 workouts

-0.65

ource

-0.64

lined

-0.64

 chalk

-0.64

halla

-0.62

 tracker

-0.62

POSITIVE LOGITS

 University

0.80

ppings

0.78

 Priv

0.76

 Sans

0.72

Marginal

0.71

University

0.69

Royal

0.69

Jur

0.67

ENS

0.67

IFA

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.