INDEX

Explanations

the definite and indefinite articles in sentences

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

is called

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

 Unterscheidung

-0.46

 quizá

-0.43

 überrascht

-0.43

 Verständnis

-0.43

 tradisi

-0.42

 wenigen

-0.41

 Ebenen

-0.41

 spørsmål

-0.41

 wielu

-0.40

 overras

-0.40

POSITIVE LOGITS

SharedCtor

0.62



0.61

“

0.59

 CALLED

0.57

(&:

0.56

‘

0.55

úgó

0.54

称为

0.54

 Called

0.54

 Camb

0.53

Activations Density 0.014%

the definite and indefinite articles in sentences

is called

No Comments

No Known Activations