INDEX

Explanations

references to individuals, specifically people with the titles "Mr." and "Dr."

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

titles before names

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

 estekak

-0.51

iecie

-0.43

 Bork

-0.42

-0.40

/\.

-0.40

 Drittan

-0.40

Spoljašnje

-0.40

columnwidth

-0.39

</tfoot>

-0.39

::~

-0.39

POSITIVE LOGITS

 newOwner

0.57

 مشين

0.56

 FontWeight

0.56

tır

0.55

 FetchType

0.53

 ciasc

0.51

Contours

0.50

ſtre

0.50

 auroit

0.50

PerformLayout

0.49

Activations Density 0.015%

references to individuals, specifically people with the titles "Mr." and "Dr."

titles before names

No Comments

No Known Activations