INDEX

Explanations

references to preferences and brand loyalty

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

personal preference and liking

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

 purpoſe

-0.43

tonsoft

-0.37

énario

-0.35

 cả

-0.35

InputTagHelper

-0.34

apimachinery

-0.34

AppRoutingModule

-0.33

URIComponent

-0.32

 viewDidLoad

-0.32

octanol

-0.32

POSITIVE LOGITS

sidemargin

0.48

 Meisters

0.47

RectangleBorder

0.46

 ddelweddau

0.46

 pinulongan

0.45

 preferido

0.45

 ergonom

0.45

 préféré

0.44

PreferredItem

0.44

 aérienne

0.43

Activations Density 0.035%

references to preferences and brand loyalty

personal preference and liking

No Comments

No Known Activations