INDEX

Explanations

proper nouns, particularly names and terms associated with people and organizations

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

Hanson or anson

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

 незавершена

-0.46

HttpDelete

-0.43

ьаж

-0.39

())));

-0.39

 consultato

-0.38

 Wiktionnaire

-0.38

 kệ

-0.37

)"),

-0.36

makeText

-0.35

 Infe

-0.34

POSITIVE LOGITS

 Vaughan

2.23

 Vaughn

1.95

 Vaugh

1.77

aughan

1.53

aughn

1.25

 Vanguard

1.09

Vanguard

1.05

 Scarborough

0.76

 Valorant

0.76

Vau

0.75

Activations Density 0.007%

proper nouns, particularly names and terms associated with people and organizations

Hanson or anson

No Comments

No Known Activations