INDEX

Explanations

terms related to mixed outcomes or varying results

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114

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

OGND

-0.94

RenderAtEndOf

-0.83

DockStyle

-0.76

GEBURTSDATUM

-0.58

 שוליים

-0.58

Биография

-0.57

queryInterface

-0.54

 Italijanski

-0.54

 пун

-0.53

\{\\

-0.53

POSITIVE LOGITS

 depending

0.44

 dependiendo

0.41

 results

0.41

 varandra

0.39

Depending

0.39

tra

0.38

δες

0.37

 divided

0.36

 divisive

0.36

Xna

0.35

Activations Density 0.043%

terms related to mixed outcomes or varying results

No Comments

No Known Activations