INDEX

Explanations

references to legal terms and familial relationships

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

various languages and specific terms

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

 greatest

-0.34

WMA

-0.33

lesia

-0.31

 ivelany

-0.31

oneg

-0.30

Cumulative

-0.30

rangle

-0.30

DeleteMapping

-0.30

ymce

-0.30

config

-0.29

POSITIVE LOGITS

 disambiguazione

0.88

 ujednoznacz

0.62

########.

0.60

 المعيارى

0.59

abestanden

0.53

Vidite

0.52

forChild

0.51

อ้างอิง

0.50

 GoogleFonts

0.50

 esternos

0.49

Activations Density 0.021%

references to legal terms and familial relationships

various languages and specific terms

No Comments

No Known Activations