INDEX

Explanations

phrases indicating sequence or hierarchy in information

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 37-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_37/width_16k/average_l0_124

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.37.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

 arşivlendi

-0.50

NameInMap

-0.41

település

-0.41

bastien

-0.40

 CrossRef

-0.37

afficheront

-0.37

Parcelize

-0.36

Huh

-0.34

kamen

-0.34

↹

-0.34

POSITIVE LOGITS

zeitige

0.67

 Gegenteil

0.63

liest

0.62

iniest

0.57

ngdoc

0.57

LookAnd

0.55

 thing

0.55

setVerticalGroup

0.53

CloseOperation

0.52

djangoproject

0.52

Activations Density 1.745%

phrases indicating sequence or hierarchy in information

No Comments

No Known Activations