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

ophon

-0.77

 subsequ

-0.75

endish

-0.74

 nickel

-0.70

oun

-0.69

quartered

-0.69

note

-0.68

oin

-0.66

 envelop

-0.66

eness

-0.64

POSITIVE LOGITS

Fla

0.76

VICE

0.74

 Profile

0.67

 Sims

0.65

IGHTS

0.65

ãĥĨ

0.65

 VIDEOS

0.63

poke

0.63

VPN

0.63

Studio

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.