INDEX

Explanations

technical terms related to model loading and saving in machine learning frameworks

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_16k/average_l0_68

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 fréqu

-0.41

 للاسماء

-0.36

 palha

-0.36

AutoresizingMask

-0.33

 bubbles

-0.33

 preguntar

-0.33

 komfort

-0.33

 ocurrir

-0.32

 visiteurs

-0.32

 asistir

-0.32

POSITIVE LOGITS

MLLoader

0.62

FormTagHelper

0.60

󠁢

0.59

ModelSerializer

0.59

 rootReducer

0.56

 AssemblyTitle

0.55

 BoxDecoration

0.53

 digress

0.50

 rested

0.49

Laure

0.49

Activations Density 0.084%

technical terms related to model loading and saving in machine learning frameworks

No Comments

No Known Activations