INDEX

Explanations

past forms of verbs and conjunctions indicating actions or states

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

states of being

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

 Bewußt

-0.37

等的

-0.36

 prédé

-0.35

€,

-0.34

✭✭

-0.34

droje

-0.33

[],

-0.33

Reserv

-0.32

 réserv

-0.32

("="

-0.32

POSITIVE LOGITS

 Administrativna

0.65

fromnode

0.62

 autorytatywna

0.58

 ProtoMessage

0.54

/**

0.52

0.51

 estekak

0.51

MemoryWarning

0.50

 otomatig

0.50

FunctionFlags

0.49

Activations Density 0.160%

past forms of verbs and conjunctions indicating actions or states

states of being

No Comments

No Known Activations