INDEX

Explanations

technical terms and specific measurements related to scientific or medical contexts

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 36-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_36/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.36.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

EA

-0.73

EO

-0.71

DCA

-0.70

AE

-0.68

OE

-0.68

IA

-0.68

OA

-0.68

AA

-0.68

QE

-0.68

KC

-0.67

POSITIVE LOGITS

 Italij

0.63

 damska

0.56

$_"

0.52

 contextLoads

0.52

 awakeFromNib

0.50

genodigd

0.50

setAccessible

0.49

 picioare

0.48

 acronym

0.48

 paž

0.48

Activations Density 0.606%

technical terms and specific measurements related to scientific or medical contexts

No Known Activations