INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

 royalties

-0.73

 Reich

-0.66

-0.65

 Hert

-0.64

 curfew

-0.62

ettings

-0.62

 broadcasts

-0.61

 Verse

-0.60

henko

-0.60

POSITIVE LOGITS

ruck

0.78

inctions

0.73

soType

0.72

igor

0.69

max

0.68

awed

0.67

ealous

0.67

abouts

0.67

assed

0.66

alling

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.