INDEX

Explanations

news sources and publications mentioned in articles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

kháu

-0.78

 abstrait

-0.71

 RIPRODUZIONE

-0.70

loài

-0.69

 dalamnya

-0.68

 Paglinawan

-0.67

$_"

-0.67

 anſ

-0.66

 bonté

-0.66

 faſt

-0.65

POSITIVE LOGITS

 estimés

0.80

 مشين

0.78

 okuyayım

0.69

ագրություններ

0.65

 arşivlendi

0.65

verwijspagina

0.64

 initComponents

0.63

KommentareTeilen

0.59

0.58

 مرئيه

0.58

Activations Density 0.229%

news sources and publications mentioned in articles

No Comments

No Known Activations