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

igmatic

-0.71

requ

-0.69

gg

-0.69

rians

-0.67

dark

-0.65

 Illum

-0.65

 Witch

-0.64

shall

-0.64

dating

-0.64

tions

-0.62

POSITIVE LOGITS

ufact

0.65

 Joint

0.63

 Nieto

0.62

 joint

0.61

 Atkinson

0.61

 periphery

0.60

Introduced

0.59

avage

0.59

 OPEC

0.58

achu

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.