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

staking

-0.84

ury

-0.67

Uncommon

-0.67

bestos

-0.66

 unden

-0.66

SPONSORED

-0.64

 arrang

-0.64

ardy

-0.63

opic

-0.62

raped

-0.61

POSITIVE LOGITS

yip

0.75

ĳå£«

0.68

 Missile

0.64

ery

0.62

mire

0.62

 Sect

0.61

Rim

0.60

 nexus

0.60

 Acceler

0.60

 Analytics

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.