INDEX

Explanations

references to vendors and companies

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

company lists

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

 flash

-0.43

isNaN

-0.41

 flashes

-0.40

 postId

-0.39

 reversible

-0.38

 poker

-0.35

 tapas

-0.35

 mild

-0.34

 pala

-0.34

 routine

-0.33

POSITIVE LOGITS

 companies

0.77

 bedrijven

0.71

 compañías

0.69

 للمعارف

0.66

companies

0.66

 yrity

0.65

 Companies

0.65

 suppliers

0.64

 şirket

0.63

 aziende

0.63

Activations Density 0.080%

references to vendors and companies

company lists

No Comments

No Known Activations