INDEX

Explanations

elements related to literary content and features

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

lists of things to be found

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

########.

-0.69

verwijspagina

-0.57

AutoScaleMode

-0.55

 Meksiku

-0.49

 BigNumberish

-0.47

GEBURTSDATUM

-0.46

 iNdEx

-0.46

positor

-0.45

epik

-0.45

 otomatig

-0.45

POSITIVE LOGITS

PreferredItem

0.48

 Exciting

0.47

 esclavos

0.45

 exciting

0.44

 BONUS

0.43

 sorpresas

0.42

Exciting

0.40

 spannende

0.39

 będ

0.39

ientôt

0.39

Activations Density 0.055%

elements related to literary content and features

lists of things to be found

No Comments

No Known Activations