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

stro

-0.65

istries

-0.65

hum

-0.64

 breakthrough

-0.64

 lessons

-0.64

apo

-0.63

ouri

-0.62

ions

-0.62

 rehab

-0.61

 collaborations

-0.61

POSITIVE LOGITS

ãĤ¨ãĥ«

0.83

Night

0.73

ãĤ¤ãĥĪ

0.70

 Night

0.69

ãĤµ

0.68

-+

0.68

ãĤ®

0.67

NPR

0.66

ãĥ´

0.65

 Chimera

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.