INDEX

Explanations

instances of observation or awareness

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

knowing and awareness

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.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

otus

-0.44

anker

-0.42

Cox

-0.41

 initComponents

-0.40

 Dona

-0.40

 fils

-0.39

onds

-0.39

led

-0.39

osh

-0.39

atorze

-0.39

POSITIVE LOGITS

 ModelExpression

0.63

 StringTokenizer

0.59

 autorytatywna

0.57

ніципалі

0.57

Rüyada

0.56

toHaveBeenCalled

0.56

 ProtoMessage

0.54

nloa

0.51

 Савезне

0.51

تقاوى

0.50

Activations Density 0.077%

instances of observation or awareness

knowing and awareness

No Comments

No Known Activations