INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

Catal

-0.72

AppData

-0.71

Rated

-0.69

 Veter

-0.68

visor

-0.68

Pir

-0.67

bid

-0.66

 externalToEVAOnly

-0.66

Phys

-0.66

Variable

-0.65

POSITIVE LOGITS

illion

0.76

adeon

0.74

 road

0.67

oola

0.67

 curfew

0.67

 roads

0.66

acia

0.66

akeru

0.65

oos

0.65

 forks

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.