INDEX

Explanations

references to classes or categories in various contexts

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

categories and classes

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_24/width_131k/average_l0_97

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.24.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

initComponents

-0.35

 Condiciones

-0.33

 contigo

-0.30

 comunale

-0.30

 Política

-0.29

 firmado

-0.29

 buatan

-0.28

 Andersson

-0.28

 halbe

-0.28

Effect

-0.27

POSITIVE LOGITS

 category

0.90

 categories

0.80

 Category

0.73

 ItemType

0.73

 kategori

0.72

 CATEGORY

0.71

 categorie

0.70

<unused51>

0.69

<unused47>

0.69

<pad>

0.69

Activations Density 0.096%

references to classes or categories in various contexts

categories and classes

No Comments

No Known Activations