INDEX

Explanations

tokens that represent specific character strings or symbols in various languages or contexts

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

spo

-0.39

AutoSize

-0.37

 JERSEY

-0.37

 KANSAS

-0.37

 قابل

-0.36

oader

-0.36

 heraus

-0.35

contentLoaded

-0.34

setAuto

-0.34

 ability

-0.33

POSITIVE LOGITS

 Myanmar

1.11

Myanmar

1.01

 Burmese

0.92

 Burma

0.88

 للاسماء

0.75

 utafitiHapana

0.69

 Aung

0.68

 intptr

0.63

 kasarigan

0.63

AnimationsModule

0.59

Activations Density 0.004%

tokens that represent specific character strings or symbols in various languages or contexts

No Comments

No Known Activations