INDEX

Explanations

words or phrases related to achievement and effectiveness

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

condition and status

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 vacía

-0.46

 légal

-0.43

 industriales

-0.43

cluye

-0.42

 académico

-0.42

 sécur

-0.41

 vacío

-0.40

 vecind

-0.40

 asegurado

-0.39

 emprego

-0.39

POSITIVE LOGITS

transQ

0.63

Parcelize

0.61

ChildScrollView

0.60

]")]

0.58

ंदीखरीदारी

0.57

 jspb

0.54

endphp

0.54

tagHelperRunner

0.54

 GenerationType

0.53

 Codable

0.53

Activations Density 0.120%

words or phrases related to achievement and effectiveness

condition and status

No Comments

No Known Activations