INDEX

Explanations

specific identifiers and proper nouns related to characters, places, and categories

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

people and places

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

</h3>

-0.32

</h2>

-0.30

 excellent

-0.29

uns

-0.27

 dedans

-0.25

↵

-0.24

 directly

-0.23

POSITIVE LOGITS

Bibitem

0.79

OGND

0.78

ſammen

0.77

 informée

0.77

 Италијани

0.77

 kohdetta

0.76

 surla

0.76

səhifə

0.74

astéroïdes

0.73

 للاسماء

0.73

Activations Density 0.156%

specific identifiers and proper nouns related to characters, places, and categories

people and places

No Comments

No Known Activations