INDEX

Explanations

specific formatting or layout attributes related to documents or systems

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 12-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_12/width_16k/average_l0_130

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.12.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

 utilice

-0.34

 skú

-0.34

 GRATU

-0.34

 romantique

-0.34

 aseg

-0.33

 spôso

-0.33

 étrangers

-0.32

 serangga

-0.32

 fémin

-0.32

 démocr

-0.32

POSITIVE LOGITS

ViewFeatures

0.77

NameInMap

0.65

ValueStyle

0.64

 ProtoMessage

0.62

hyrchwyd

0.60

Personendaten

0.60

 يتيمه

0.60

WireFormatLite

0.59

хьтан

0.58

adaptiveStyles

0.57

Activations Density 0.033%

specific formatting or layout attributes related to documents or systems

No Comments

No Known Activations