INDEX

Explanations

the presence of end-of-turn markers in dialogue or conversation

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

Imperial, Lancaster, Staf

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

OGND

-0.70

 kasarigan

-0.69

 الرياضيه

-0.63

transQ

-0.62

 يتيمه

-0.60

ConstraintMaker

-0.57

kháu

-0.54

 المعيارى

-0.54

xase

-0.53

Бахар

-0.52

POSITIVE LOGITS

zä

0.34

 snippetHide

0.33

}{|

0.32

 Imperial

0.32

 Lancaster

0.31

 Staf

0.31

 брен

0.31

 buzz

0.30

 backdrop

0.30

 επικ

0.30

Activations Density 0.153%

the presence of end-of-turn markers in dialogue or conversation

Imperial, Lancaster, Staf

No Comments

No Known Activations