INDEX

Explanations

references to hierarchical structures and classifications

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

 mergeFrom

-0.39

 reger

-0.37

jącej

-0.36

orianCalendar

-0.35

Hentet

-0.34

 sucre

-0.34

sejarah

-0.34

 regiões

-0.34

 dager

-0.34

Seu

-0.33

POSITIVE LOGITS

 linkovi

0.49

GEBURTSDATUM

0.48

fromCharCode

0.45

matchCondition

0.44

 kasarigan

0.43

 crí

0.43

getLevel

0.43

 level

0.43

lvl

0.43

SharedDtor

0.42

Activations Density 0.120%

references to hierarchical structures and classifications

No Comments

No Known Activations