INDEX

Explanations

expressions of affection and relationships

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

is followed by descriptive words

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

Economia

-0.42

)))))

-0.42

 referenties

-0.34

RegressionTest

-0.34

))))

-0.33

 Bernstein

-0.32

]]);

-0.32

 Backman

-0.32

 Nowak

-0.32

 (“

-0.32

POSITIVE LOGITS

SequentialGroup

0.68

 برانيه

0.66

QMetaType

0.65

ésultats

0.65

niſſe

0.63

appré

0.61

 للمعارف

0.60

 ſei

0.60

ſchaft

0.59

нгред

0.59

Activations Density 0.070%

expressions of affection and relationships

is followed by descriptive words

No Comments

No Known Activations