INDEX

Explanations

proper nouns, particularly names of celebrities and public figures

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

people's names

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

antMatchers

-0.39

 الحره

-0.38

phosa

-0.38

Frauen

-0.36

 guarn

-0.36

Mog

-0.35

UIControlState

-0.35

TestingModule

-0.34

 ẩn

-0.33

frauen

-0.33

POSITIVE LOGITS

 wikipagina

0.63

 Roskov

0.59

OGND

0.56

PreferredItem

0.54

 ehemalige

0.52

Климаты

0.49

ArrowToggle

0.46

省市镇

0.46

Дереккөздер

0.45

يكب

0.45

Activations Density 0.009%

proper nouns, particularly names of celebrities and public figures

people's names

No Comments

No Known Activations