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

-+-+

-0.69

 Corps

-0.67

 Sins

-0.66

Ø¯

-0.66

ãĤĮ

-0.66

arthed

-0.65

 Mong

-0.65

SD

-0.65

OUP

-0.63

 Instruct

-0.60

POSITIVE LOGITS

rix

0.71

 checkout

0.65

 Outlook

0.63

 insol

0.62

anqu

0.62

icia

0.61

 Kimber

0.61

htaking

0.60

hips

0.60

eli

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.