INDEX

Explanations

personal pronouns, particularly references to the second person "you" and possessive forms like "your."

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 13-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_13/width_65k/average_l0_75

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.13.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

 AssemblyTitle

-0.54

Ne

-0.50

 giro

-0.49

十分

-0.46

 Gros

-0.43

gesellschaft

-0.42

 denn

-0.42

 Particularly

-0.42

 gelo

-0.41

全文

-0.40

POSITIVE LOGITS

your

0.95

 your

0.89

 propOrder

0.83

YOUR

0.80

 yourself

0.74

 noDo

0.74

 Мексичка

0.74

Your

0.73

 YOUR

0.73

votre

0.72

Activations Density 0.061%

personal pronouns, particularly references to the second person "you" and possessive forms like "your."

No Known Activations