INDEX

Explanations

complex relationships and exceptions within textual descriptions

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

linking common words

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

inheritDoc

-0.35

Voltaje

-0.34

createStatement

-0.34

 publicitaires

-0.33

 ordini

-0.33

 couteau

-0.31

 lær

-0.31

TabView

-0.31

 commandement

-0.30

 chrétien

-0.30

POSITIVE LOGITS

astéroïdes

0.81

 estekak

0.77

 nakalista

0.74

 հղումներ

0.71

 الحره

0.67

NameInMap

0.65

 للمعارف

0.64

Билгалдахарш

0.63

 cherchés

0.63

ьаж

0.62

Activations Density 0.043%

complex relationships and exceptions within textual descriptions

linking common words

No Comments

No Known Activations