INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 legally

-0.73

 voluntarily

-0.66

 advertisers

-0.66

 administr

-0.65

 artificially

-0.64

 willfully

-0.64

 sustained

-0.64

 Dash

-0.61

RED

-0.61

 resc

-0.60

POSITIVE LOGITS

credit

1.10

SPONSORED

1.01

Weapon

0.95

Greek

0.95

Asset

0.92

Filename

0.90

soType

0.87

Aust

0.87

Mass

0.87

Credit

0.87

Activations Density 0.000%

No Known Activations

This feature has no known activations.