INDEX

Explanations

phrases indicating first-person statements or opinions

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

please explain

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

TagMode

-0.38

EndGlobalSection

-0.38

zczegól

-0.36

 AppCompatTheme

-0.36

ViewFeatures

-0.35

 components

-0.34

 שוליים

-0.33

/*!

-0.32

ta

-0.32

prüche

-0.32

POSITIVE LOGITS

tvguidetime

0.52

 propOrder

0.52

 ویکی‌آمباردا

0.43

 Learners

0.43

教學

0.42

 Tunnel

0.42

<?

0.41

namese

0.41

 Inoc

0.41

 Educator

0.40

Activations Density 0.007%

phrases indicating first-person statements or opinions

please explain

No Comments

No Known Activations