INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

ãĥīãĥ©ãĤ´ãĥ³

-0.75

nia

-0.70

ICES

-0.69

iguous

-0.67

 gems

-0.67

 Samoa

-0.65

ILA

-0.65

Gin

-0.65

 Genie

-0.64

 ener

-0.64

POSITIVE LOGITS

rection

0.72

 Breach

0.69

morrow

0.67

ereo

0.65

claw

0.65

ongyang

0.65

hander

0.64

smith

0.63

olphin

0.62

kn

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.