INDEX

Explanations

HTML comment tags and their contents

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

configuration attributes

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

 Autorizaciones

-0.63

 فريبيس

-0.54

endphp

-0.52

 snippetHide

-0.52

 kasarigan

-0.51

 sanctuary

-0.49

 CanadaChoose

-0.49

tangentMode

-0.48

 '\\;'

-0.47

Bif

-0.47

POSITIVE LOGITS

 näin

0.39

 justement

0.37

 thôi

0.37

These

0.35

AndEndTag

0.34

 enorme

0.33

 wiederum

0.32

 these

0.32

 wundersch

0.32

 szóci

0.32

Activations Density 0.010%

HTML comment tags and their contents

configuration attributes

No Comments

No Known Activations