INDEX

Explanations

references to the concept of work and effort across various contexts

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-mlp/layer_6/width_131k/average_l0_72

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.6.hook_mlp_out

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



-0.62

 work

-0.57

 disambiguazione

-0.44

work

-0.43

 работы

-0.43

Work

-0.40

 WORK

-0.39

 Work

-0.39

Pranala

-0.39

 Arbeiten

-0.38

POSITIVE LOGITS

Personendaten

0.63

 Biôgrafia

0.54

 الحره

0.51

原始内容存档于

0.50

 Савезне

0.49

Obrázky

0.47

 '\\;'

0.47

inaison

0.47

 CreateTagHelper

0.47

 ProtoMessage

0.47

Activations Density 0.005%

references to the concept of work and effort across various contexts

No Comments

No Known Activations