INDEX

Explanations

references to variations and versions of objects or concepts

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

versions, counterparts, and ones

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

 surla

-0.58

TagMode

-0.58

-0.57

 OFDb

-0.55

VersionUID

-0.53

timedia

-0.53

awaiter

-0.51

Hozzáférés

-0.50

TintMode

-0.50

 ujednoznacz

-0.49

POSITIVE LOGITS

 helft

0.60

 counterpart

0.55

 mijne

0.54

 Hälfte

0.48

 version

0.47

 Variante

0.47

 zijne

0.46

 löyty

0.45

 metade

0.45

 nôtre

0.43

Activations Density 0.061%

references to variations and versions of objects or concepts

versions, counterparts, and ones

No Comments

No Known Activations