INDEX

Explanations

references to names, likely identifying specific individuals or entities

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

kman, maru, dejuren, GringoStar

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

 autorytatywna

-0.85

MigrationBuilder

-0.82

oa̍t

-0.80

 الرياضيه

-0.78

 незавершена

-0.77

Diweddarwch

-0.77

ValueStyle

-0.76

 kasarigan

-0.75

ConstraintMaker

-0.75

səhifə

-0.74

POSITIVE LOGITS

 wrote

0.40

 CURIAM

0.35

0.32

 himself

0.32

 corrected

0.31

’

0.31

 personalmente

0.31

Activations Density 0.046%

references to names, likely identifying specific individuals or entities

kman, maru, dejuren, GringoStar

No Comments

No Known Activations