INDEX

Explanations

references to evil and malevolent characters or forces

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 0-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_0/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.0.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Volkes

-0.51

 Confederation

-0.50

 perfiles

-0.50

 nahilalakip

-0.49

ufficient

-0.49

#+#

-0.48

 plana

-0.48

IUrlHelper

-0.48

 profilo

-0.47

AnchorTagHelper

-0.46

POSITIVE LOGITS

 robots

0.66

 robotic

0.56

 robotics

0.54

ſelf

0.50

 robot

0.49

WireFormatLite

0.47

 armor

0.47

ski

0.47

 armour

0.46

 Robotics

0.46

Activations Density 0.146%

references to evil and malevolent characters or forces

No Comments

No Known Activations