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

 Towns

-0.75

ridor

-0.75

tics

-0.72

thumbnails

-0.72

elsen

-0.72

redd

-0.71

orable

-0.71

outube

-0.70

oids

-0.70

izont

-0.69

POSITIVE LOGITS

—

1.31

—"

1.05

"—

1.02

)—

1.01

,—

0.92

--

0.91

.—

0.80

—-

0.79

Write

0.69

–

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.