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

 Flavoring

-0.82

 shone

-0.72

 flaws

-0.70

CVE

-0.68

 waived

-0.68

ESA

-0.65

 waivers

-0.65

 rocked

-0.64

 shines

-0.64

 surpassed

-0.64

POSITIVE LOGITS

uez

0.80

 Gujar

0.70

Interstitial

0.67

anim

0.67

monary

0.66

inton

0.66

 Arrows

0.65

 mosquit

0.65

Eag

0.64

Zed

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.