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

 clad

-0.75

enium

-0.74

 ribbon

-0.73

fle

-0.67

 handsome

-0.66

 corrosion

-0.64

 ornament

-0.63

 furnished

-0.63

 comple

-0.63

 nickel

-0.63

POSITIVE LOGITS

ba

0.87

geist

0.77

SPA

0.74

 Mate

0.70

 Cance

0.69

MU

0.68

ESA

0.66

vana

0.66

 Related

0.66

 Advocate

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.