INDEX

Explanations

personal names and their associated positive attributes

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

David sometimes, writing

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

 anonymously

-0.40

 Bazar

-0.38

ENSIS

-0.38

withIdentifier

-0.38

 AssemblyProduct

-0.37

 Breakers

-0.37

 Anonymous

-0.36

 Popular

-0.36

ktır

-0.35

Pinus

-0.35

POSITIVE LOGITS

 مرئيه

0.56

 pinulongan

0.56

WriteTagHelper

0.48

 שוליים

0.47

GeneratedMessage

0.46

 DevComponents

0.45

deutig

0.44

Efq

0.44

 Administrativna

0.43

 мәкал

0.43

Activations Density 0.004%

personal names and their associated positive attributes

David sometimes, writing

No Comments

No Known Activations