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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Asia

-0.77

Ther

-0.75

oret

-0.71

rooms

-0.71

Nature

-0.71

own

-0.69

 Boyle

-0.68

Talk

-0.68

 Notting

-0.67

Fac

-0.66

POSITIVE LOGITS

 destro

0.78

 straps

0.76

 strap

0.75

0.74

 cord

0.73

sung

0.71

jug

0.71

 deducted

0.71

xus

0.69

scl

0.68

Activations Density 0.000%

No Known Activations

This feature has no known activations.