INDEX

Explanations

phrases indicating negation or insufficiency

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

even phrases

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

出版年

-0.45

 '\\;'

-0.41

лтемелер

-0.40

 الرياضيه

-0.40

 Dauer

-0.40

 aérienne

-0.39

 aéreo

-0.39

 geslacht

-0.38

 Geister

-0.38

 invers

-0.38

POSITIVE LOGITS

CppMethod

0.52

rungsseite

0.48

featureID

0.46

RRect

0.45

Worse

0.45

⇡

0.44

worse

0.44

ionView

0.44

msgTypes

0.43

 IActionResult

0.42

Activations Density 0.089%

phrases indicating negation or insufficiency

even phrases

No Comments

No Known Activations