INDEX

Explanations

expressions of gratitude or phrases indicating a positive recommendation

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

multilingual conjunctions and evaluative terms

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

AutoField

-0.40

 eleições

-0.40

 InputDecoration

-0.38

rrggbb

-0.38

 establecidos

-0.38

 península

-0.38

 transférez

-0.37

 conformidad

-0.36

 '\\;'

-0.36

 prisión

-0.35

POSITIVE LOGITS

helpful

0.52

ByPrimaryKey

0.48

awtextra

0.47

accro

0.46

 Winaray

0.46

باردا

0.46

 helpful

0.45

 graphical

0.44

 unhelpful

0.44

 visuals

0.43

Activations Density 0.034%

expressions of gratitude or phrases indicating a positive recommendation

multilingual conjunctions and evaluative terms

No Comments

No Known Activations