INDEX

Explanations

phrases indicating the act of coming to or addressing a topic or subject

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

when it comes to

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

 للمعارف

-0.62

GEBURTSDATUM

-0.59

 Мексичка

-0.55



-0.52

 يتيمه

-0.52

isContained

-0.49

setVerticalGroup

-0.49

UrlResolution

-0.49

findpost

-0.47

 arşivlendi

-0.47

POSITIVE LOGITS

 letteratura

0.42

 rivalry

0.42

 psychologically

0.40

rimônio

0.38

dhism

0.38

 scalability

0.38

HttpEntity

0.38

openhauer

0.38

xtype

0.38

 Bedroom

0.37

Activations Density 0.014%

phrases indicating the act of coming to or addressing a topic or subject

when it comes to

No Comments

No Known Activations