INDEX

Explanations

references to image files, particularly icons and their attributes

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

icon and symbol

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

EnableWeb

-0.37

 âgées

-0.37

 koszt

-0.37

crollView

-0.36

GraphicsUnit

-0.36

 stara

-0.35

 pressure

-0.35

 Wilk

-0.35

 tramo

-0.35

 estrés

-0.34

POSITIVE LOGITS

 فريبيس

0.79

 icons

0.73

 symbols

0.66

 icon

0.66

 defaultstate

0.66

ListItemIcon

0.56

 &___

0.54

 initComponents

0.54

ymbol

0.54

Tikang

0.53

Activations Density 0.017%

references to image files, particularly icons and their attributes

icon and symbol

No Comments

No Known Activations