INDEX

Explanations

references to various "premier" entities, likely in the context of sports or notable categories

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

Premier League, Primera División

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

Bill

-0.56

 Беларусь

-0.54

bill

-0.53

debt

-0.52

Debt

-0.52

XOXO

-0.52

 Bill

-0.50

Krank

-0.50

Fußball

-0.50

POSITIVE LOGITS

 PREMIER

0.90

 Premier

0.88

Premier

0.82

 premier

0.75

premier

0.73

PREMI

0.63

 Premiere

0.61

 surla

0.60

 premiers

0.59

 adquis

0.57

Activations Density 0.003%

references to various "premier" entities, likely in the context of sports or notable categories

Premier League, Primera División

No Comments

No Known Activations