INDEX

Explanations

mathematical comparisons and references to specific units of measurement

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

less than, dot com

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

Rptr

-0.48

CDCl

-0.45

OGND

-0.44

UserScript

-0.43

IContainer

-0.43

AnimationsModule

-0.43

ChildScrollView

-0.42

">//

-0.41

UAWEI

-0.41

 iconLine

-0.41

POSITIVE LOGITS

 يتيمه

0.76

 minus

0.61

 pyplot

0.56

 PRODU

0.55

archiviato

0.55

 Giovan

0.54

 kilogram

0.53

 plus

0.53

entgen

0.53

 rectangle

0.52

Activations Density 0.311%

mathematical comparisons and references to specific units of measurement

less than, dot com

No Comments

No Known Activations