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

 Landing

-0.69

ERY

-0.69

MI

-0.67

Wars

-0.63

Wilson

-0.62

 Recovery

-0.62

LOCK

-0.61

 Asylum

-0.61

Associated

-0.61

mart

-0.60

POSITIVE LOGITS

tes

0.86

ihu

0.84

 nodd

0.83

herical

0.78

nces

0.78

ptin

0.76

risome

0.76

zin

0.74

 challeng

0.73

 destro

0.72

Activations Density 0.000%

No Known Activations

This feature has no known activations.