INDEX

Explanations

technical terms, variable names, and code-related constructs

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

Spanish/German words

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

хьтан

-0.88

 autorytatywna

-0.68

numerusform

-0.67

Autoritní

-0.64

OGND

-0.64

ronpa

-0.61

 CURIAM

-0.61

VideoCapture

-0.60

hyrchwyd

-0.59

 betweenstory

-0.59

POSITIVE LOGITS

.*")]

0.34

uxxxx

0.34

 cuchar

0.33

 Flügel

0.32

 ofreci

0.30

 criterios

0.30

 desarroll

0.30

 usarlo

0.29

 cuotas

0.29

("")]

0.29

Activations Density 0.163%

technical terms, variable names, and code-related constructs

Spanish/German words

No Comments

No Known Activations