INDEX

Explanations

occurrences of abbreviations and related notation in the text

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 AssemblyTitle

-0.37

 utilisateurs

-0.37

PreInfinity

-0.36

 integrado

-0.34

 officiers

-0.34

 AssemblyCulture

-0.34

 users

-0.33

setPower

-0.33

 getUsers

-0.33

 bénéficiaire

-0.33

POSITIVE LOGITS

 Италијани

0.73



0.69

 يتيمه

0.59

хьтан

0.59

 незавершена

0.57

rawDesc

0.55

 Italijanski

0.54

CppMethod

0.54

verwijspagina

0.52

 ویکی‌پدی

0.52

Activations Density 0.068%

occurrences of abbreviations and related notation in the text

No Comments

No Known Activations