INDEX

Explanations

phrases related to definitions and their contextual usages

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

 disambiguazione

-0.53

setVerticalGroup

-0.47

:+:

-0.44

cristo

-0.43

hyrchwyd

-0.42

 Exactos

-0.41

 autorytatywna

-0.41

RTEX

-0.41

HasForeignKey

-0.41

Chham

-0.41

POSITIVE LOGITS

 distance

0.49

culated

0.46

 GenerationType

0.39

XtraBars

0.37

distance

0.36

 DISTANCE

0.36

それ

0.36

dafx

0.35

 Distance

0.35

рез

0.35

Activations Density 0.065%

phrases related to definitions and their contextual usages

No Comments

No Known Activations