INDEX

Explanations

references to applications or software tools

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

sizeCache

-0.66

 للمعارف

-0.66

MigrationBuilder

-0.63

parsedMessage

-0.60

 miniaturka

-0.60

featureID

-0.59

 ***!

-0.58

Espèce

-0.57

 indígen

-0.57

 Comprometido

-0.57

POSITIVE LOGITS

vernote

0.48

 organize

0.44

 organise

0.41

 desk

0.41

 organizing

0.39

yak

0.38

 Desk

0.37

 workflow

0.37

Organize

0.36

 organising

0.36

Activations Density 0.033%

references to applications or software tools

No Comments

No Known Activations