INDEX

Explanations

punctuation marks that indicate sentence structure or separation

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

text segmentation punctuation

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

 étoient

-0.48

 avoient

-0.45

 enfans

-0.44

 feroit

-0.40

 maquin

-0.34

 emprend

-0.34

 cucharadita

-0.33

Specyfikacja

-0.32

Notae

-0.31

 bienfaits

-0.31

POSITIVE LOGITS

Clik

0.65

rectal

0.63

 שוליים

0.63

recated

0.59

Vidite

0.58

astéro

0.57

cesis

0.57

انيف

0.56

onalds

0.56

enery

0.56

Activations Density 0.127%

punctuation marks that indicate sentence structure or separation

text segmentation punctuation

No Comments

No Known Activations