INDEX

Explanations

punctuation marks indicating pauses or separate thoughts in dialogue

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

reporting speech

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

StoryboardSegue

-0.60

 autorytatywna

-0.60

 nahilalakip

-0.60

fjspx

-0.59

Personensuche

-0.57

 ویکی‌آمباردا

-0.54

VIAF

-0.52

ftagPool

-0.52

المناصب

-0.52

 &___

-0.51

POSITIVE LOGITS

 said

0.57

 says

0.54

 according

0.52

 ifølge

0.51

 según

0.46

 enligt

0.40

Empereur

0.40

 selon

0.40

 experts

0.39

 volgens

0.39

Activations Density 0.041%

punctuation marks indicating pauses or separate thoughts in dialogue

reporting speech

No Comments

No Known Activations