INDEX

Explanations

phrases that suggest significant importance or critical roles in various contexts

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

possibility or potential effect

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

 commandement

-0.43

expandindo

-0.40

 basically

-0.40

XmlIgnore

-0.39

 docteur

-0.38

habitude

-0.37

ars

-0.37

Rohy

-0.35

gula

-0.35

 AssemblyCompany

-0.35

POSITIVE LOGITS

 kanskje

0.69

 måske

0.68

anskje

0.66

 évent

0.66

 might

0.66

 eventuali

0.66

 möglicherweise

0.64

 kanske

0.62

 eventuell

0.60

อาจ

0.59

Activations Density 0.371%

phrases that suggest significant importance or critical roles in various contexts

possibility or potential effect

No Comments

No Known Activations