INDEX

Explanations

code that defines or creates new functions or 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

transQ

-0.53

 StringTokenizer

-0.47

 IActionResult

-0.47

ByVersion

-0.46

 فريبيس

-0.45

 ModelExpression

-0.44

 treason

-0.43

jsdelivr

-0.43

 rospy

-0.43

=$("#

-0.43

POSITIVE LOGITS

setVerticalGroup

0.81

 initComponents

0.62

writerow

0.43

 Numerade

0.42

 GreatSchools

0.41

 gesteld

0.41

PropertyGroup

0.40

 AssemblyTitle

0.40

 varandra

0.40

Autoritní

0.40

Activations Density 0.000%

code that defines or creates new functions or components

No Comments

No Known Activations