INDEX

Explanations

image file formats and related attributes in code snippets

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 للمعارف

-0.86

bootstrapcdn

-0.79

 Infórmanos

-0.78

 يتيمه

-0.78

 kasarigan

-0.76

ſſung

-0.72

oa̍t

-0.72

 Мексичка

-0.71

########.

-0.70

<unused14>

-0.69

POSITIVE LOGITS

 jsPsych

0.29

BindingSource

0.26

0.24

 écrivain

0.23

stellen

0.23

ưng

0.23

ůž

0.23

 Schild

0.23

 Personalis

0.23

 implements

0.23

Activations Density 0.003%

image file formats and related attributes in code snippets

No Comments

No Known Activations