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

 Mong

-0.77

.–

-0.73

EngineDebug

-0.69

 CrossRef

-0.68

 Napoleon

-0.67

 Osaka

-0.67

 Axis

-0.66

 Sloan

-0.65

 Pros

-0.63

 Mongol

-0.62

POSITIVE LOGITS

hold

1.39

 grabs

1.14

ãģ®å

0.86

ht

0.76

itance

0.76

§

0.75

holding

0.75

aches

0.70

Ã°

0.70

aim

0.69

Activations Density 0.000%

No Known Activations

This feature has no known activations.