INDEX

Explanations

references to numerical identifiers, specifically related to academic papers and their citations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.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

 doPost

-0.66

ốm

-0.63

érêt

-0.61

Ārējās

-0.60

ToDecimal

-0.60

 tartalomajánló

-0.59

 NgModule

-0.58

 kín

-0.58

 serializers

-0.57

 Vieira

-0.56

POSITIVE LOGITS

 BoxDecoration

0.68

UIControlState

0.66

BoxShadow

0.63

 مشين

0.62

CloseOperation

0.61

 الدولى

0.58

بوابة

0.57

+#+

0.53

ppelin

0.52

 بيها

0.52

Activations Density 0.043%

references to numerical identifiers, specifically related to academic papers and their citations

No Comments

No Known Activations