INDEX

Explanations

phrases expressing requests or desires involving learning or obtaining information

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 27-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_27/width_16k/average_l0_118

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.27.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

 méri

-0.56

 ninguno

-0.50

 étrangère

-0.49

 recientemente

-0.48

 lumineuse

-0.48

 descans

-0.48

 étranger

-0.48

 réus

-0.48

 entière

-0.47

 moldura

-0.47

POSITIVE LOGITS

WriteTagHelper

0.72

üyada

0.54

be

0.51

 otomatig

0.51

日閲覧

0.50

GIVEREF

0.50

gwt

0.48

----</

0.46

 Sitemap

0.45

 getEmail

0.44

Activations Density 0.095%

phrases expressing requests or desires involving learning or obtaining information

No Comments

No Known Activations