INDEX

Explanations

affirmative directives and expressions of intent

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

esistenza

-0.46

 kasarigan

-0.43

withOpacity

-0.42

 Wikiseite

-0.42

tonode

-0.42

paravant

-0.41

Ecotoxicity

-0.41

 Commentaires

-0.40

 Arxivat

-0.40

tableFuture

-0.40

POSITIVE LOGITS

Let

0.63

Vamos

0.57

Давайте

0.57

 Vamos

0.56

Let

0.56

 Lets

0.54

Lets

0.52

let

0.52

 للاسماء

0.50

 vamos

0.50

Activations Density 0.018%

affirmative directives and expressions of intent

No Comments

No Known Activations