INDEX

Explanations

phrases related to security measures and oversight

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

"guard" or "supervision"

np_acts-logits-general · gemini-2.0-flash

supervision and guarding

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.18.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

 autorytatywna

-0.70

 unknownFields

-0.58

 onItemClick

-0.55

 nahilalakip

-0.54

fjspx

-0.54

 CascadeType

-0.53

InvalidProtocol

-0.52

UpInside

-0.49

AddTagHelper

-0.48

 isKindOfClass

-0.47

POSITIVE LOGITS

 monitoring

1.39

 supervision

1.32

 surveillance

1.28

 guarding

1.28

 monitored

1.25

 guard

1.22

 Monitoring

1.21

 patrols

1.20

 patrolling

1.19

 guarded

1.16

Activations Density 0.191%

phrases related to security measures and oversight

"guard" or "supervision"

supervision and guarding

No Comments

No Known Activations