INDEX

Explanations

references to data, figures, and results in scientific or technical documents

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

citations or figures

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

تقاوى

-0.46

 мәкал

-0.43

Personendaten

-0.41

⅘

-0.41

SharedCtor

-0.41

rep

-0.40

tvguidetime

-0.40

nodoc

-0.40

 ويكيميديا

-0.38

ʖ

-0.38

POSITIVE LOGITS

 Chapter

0.45

 vocales

0.43

 chapter

0.43

แดง

0.42

 '\\;'

0.42

ApiModel

0.41

getExtras

0.41

 pesca

0.41

 étranger

0.41

 langs

0.40

Activations Density 0.102%

references to data, figures, and results in scientific or technical documents

citations or figures

No Comments

No Known Activations