INDEX

Explanations

conjunctions, particularly "and" and "or."

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

enumi

-0.43

jenigen

-0.42

antaranya

-0.40

harapkan

-0.40

 möge

-0.38

 Versorgung

-0.37

 falsche

-0.37

 izman

-0.37

 wobei

-0.37

 programu

-0.36

POSITIVE LOGITS

getOr

0.65

 ivelany

0.63

setVerticalGroup

0.57

OGND

0.57

 فريبيس

0.54

 oprot

0.53

 estekak

0.53

ActivityCompat

0.52

Transkript

0.52

 Nand

0.50

Activations Density 0.055%

conjunctions, particularly "and" and "or."

No Comments

No Known Activations