INDEX

Explanations

references to processes or actions related to suppression

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

Comes before "planted", "ants", "orters", "lying", or "ressive"

np_acts-logits-general · gemini-2.0-flash

supplanted, supplanting, supporters, suppressing

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 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

Naissance

-0.68

/***/

-0.66

veness

-0.60

 Angele

-0.52

triangleq

-0.51

antiation

-0.49

 fVar

-0.49

 Vikipedi

-0.48

pable

-0.48

밤

-0.47

POSITIVE LOGITS

msgSender

0.67



0.64

oredCriteria

0.62

таратура

0.61

})*/

0.59

HtmlAttribute

0.59

 kaarangay

0.58

httphttps

0.57

Personensuche

0.54

Viited

0.52

Activations Density 0.177%

references to processes or actions related to suppression

Comes before "planted", "ants", "orters", "lying", or "ressive"

supplanted, supplanting, supporters, suppressing

No Comments

No Known Activations