INDEX

Explanations

references to individuals or roles within a narrative context

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

names of people

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

دانشنامهٔ

-0.65

BufferException

-0.57

 snippetHide

-0.53

">//

-0.49

 cherchés

-0.49

AccessorTable

-0.49

asteroide

-0.48

 ویکی‌پدی

-0.47

تقاوى

-0.47

 otomatig

-0.46

POSITIVE LOGITS

 Stimme

0.39

tagext

0.39

 kohdetta

0.37

 imprensa

0.37

 Soldaten

0.35

 personalidade

0.35

!*\

0.34

 CURIAM

0.34

 opinião

0.34

 Einwohner

0.33

Activations Density 0.017%

references to individuals or roles within a narrative context

names of people

No Comments

No Known Activations