INDEX

Explanations

references to family relationships and dynamics

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

family members

np_acts-logits-general · gemini-2.0-flash

parents and family members

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

Datuak

-0.51

 kasarigan

-0.49

posables

-0.42

ſelves

-0.42

 éco

-0.41

 useAppContext

-0.41

Referencies

-0.41

 Gegenteil

-0.40

Сек

-0.40

 Gosling

-0.40

POSITIVE LOGITS

 grandparents

0.52

 """",

0.48

awtextra

0.47

 grandpa

0.45

gridy

0.44

 aunts

0.44

 family

0.44

ReusableCell

0.43

 grandfather

0.41

 parents

0.40

Activations Density 0.155%

references to family relationships and dynamics

family members

parents and family members

No Comments

No Known Activations