INDEX

Explanations

prominent names and titles, particularly in academic and professional contexts

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 kasarigan

-0.94

 AssemblyCulture

-0.88

 المعيارى

-0.85

astify

-0.83

ロウィン

-0.81

OGND

-0.79

 autorytatywna

-0.76

 ModelExpression

-0.75

 <=",

-0.73

 BoxFit

-0.73

POSITIVE LOGITS

städ

0.33

0.32

 John

0.31

Bagaimana

0.31

上线

0.30

 Loeb

0.30

 corred

0.30

Activations Density 0.308%

prominent names and titles, particularly in academic and professional contexts

No Comments

No Known Activations