INDEX

Explanations

references to individuals and their professional roles or backgrounds

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

former positions

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

 murni

-0.39

 manzanas

-0.36

 Artículos

-0.36

 kış

-0.34

 jälkeen

-0.33

 ciąży

-0.33

 comodidad

-0.33

 permitiendo

-0.33

 Gegenteil

-0.33

 våre

-0.33

POSITIVE LOGITS

enderror

0.69

 CreateTagHelper

0.68

الدراسه

0.67

MigrationBuilder

0.65

AndEndTag

0.63

:✨

0.63

WriteAttribute

0.62

 Administrativna

0.62

 оригіналу

0.60

 invokingState

0.59

Activations Density 0.016%

references to individuals and their professional roles or backgrounds

former positions

No Comments

No Known Activations