INDEX

Explanations

terms and details related to physical attributes and components

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

цездатний

-0.83

awaiter

-0.73

PerformLayout

-0.73

HostException

-0.73

InstrumentedTest

-0.67

HasForeignKey

-0.64

InjectAttribute

-0.63

 snippetHide

-0.61

 OFDb

-0.59

styleable

-0.59

POSITIVE LOGITS

uttgart

0.33

 whilst

0.32

bindet

0.32

élev

0.31

setContentView

0.30

Meskipun

0.29

Ārējās

0.29

 natin

0.28

Whilst

0.28

 fizer

0.27

Activations Density 0.057%

terms and details related to physical attributes and components

No Comments

No Known Activations