INDEX

Explanations

references to the New Year

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

Sean followed by surname

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 16-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_16/width_65k/average_l0_128

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.16.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

 Figue

-0.71

msgSender

-0.69

_{

-0.68

orithmic

-0.66

 Timor

-0.66

]_{

-0.65

 Valenzuela

-0.65

 Peña

-0.62

 utafitiHapana

-0.62

PathVariable

-0.60

POSITIVE LOGITS

msup

1.07

Nea

0.82

Portail

0.80

EndInit

0.78

сылкі

0.78

 Monfieur

0.71

 препратки

0.71

[][]

0.71

ChildScrollView

0.71

 Hiram

0.70

Activations Density 0.025%

references to the New Year

Sean followed by surname

No Comments

No Known Activations