INDEX

Explanations

concepts related to mathematical structures and families of objects within a theoretical framework

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

parametric family, deformations, interpolates

np_acts-logits-general · gemini-2.5-flash-lite

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 nega

-0.49

negative

-0.47

neg

-0.46

Neg

-0.45

 Negative

-0.43

 Bioaccumulative

-0.43

Negative

-0.43

 TextAppearance

-0.42

 negative

-0.42

 комби

-0.42

POSITIVE LOGITS

DockStyle

0.47

 Anlass

0.42

 يتيمه

0.42

 ModelExpression

0.42

errHandler

0.41

AndEndTag

0.41

intios

0.40

tanleria

0.40

figsize

0.40

astéroïdes

0.39

Activations Density 0.402%

concepts related to mathematical structures and families of objects within a theoretical framework

parametric family, deformations, interpolates

No Comments

No Known Activations