INDEX

Explanations

proper nouns, specifically names of people and organizations

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

authors and researchers

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

 inmig

-0.38

Географи

-0.36

SerializedName

-0.35

Билгалдахарш

-0.34

})));

-0.34

Controllo

-0.34

 Sedangkan

-0.34

 Manusia

-0.34

ortunadamente

-0.33

 useRef

-0.33

POSITIVE LOGITS

 queſto

0.55

 оригіналу

0.54

expandindo

0.52

ᅝ

0.52

๣

0.50

Насе

0.50

Ӕ

0.50

raulic

0.49

 Pauli

0.49

tius

0.49

Activations Density 0.061%

proper nouns, specifically names of people and organizations

authors and researchers

No Comments

No Known Activations