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

iliated

-0.66

 thereof

-0.64

 wards

-0.63

eto

-0.61

leans

-0.61

 effected

-0.60

ean

-0.59

 minim

-0.59

 Fundamental

-0.58

utations

-0.58

POSITIVE LOGITS

WARN

0.95

resa

0.84

Zup

0.81

aceae

0.76

NCT

0.75

 Jagu

0.74

amia

0.72

undy

0.72

 Petra

0.71

MSN

0.71

Activations Density 0.000%

No Known Activations

This feature has no known activations.