INDEX

Explanations

proper names of significant individuals and entities

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

famous historical figures

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

 gangen

-0.49

gemaakt

-0.42

appartamento

-0.42

impianto

-0.41

übt

-0.40

occasione

-0.40

før

-0.39

 redonda

-0.38

 übrig

-0.38

 bordados

-0.38

POSITIVE LOGITS

 المعيارى

0.61

 cherchés

0.58

 للمعارف

0.56

 فريبيس

0.53

CppCodeGen

0.52

ValueStyle

0.51

RegistryLite

0.51

Diwedd

0.51

EndProject

0.50

windowFixed

0.48

Activations Density 0.047%

proper names of significant individuals and entities

famous historical figures

No Comments

No Known Activations