INDEX

Explanations

numeric values or monetary amounts

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

lists ending with "and"

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

 للاسماء

-0.62

TagMode

-0.57

 kasarigan

-0.56

 חיצוניים

-0.56

########.

-0.56

LookAnd

-0.53

 surla

-0.50

DispatchToProps

-0.50

majánló

-0.49

StoryboardSegue

-0.47

POSITIVE LOGITS

 allemaal

0.42

reszcie

0.37

 lastly

0.37

 dAtA

0.36

Ад

0.34

ViewImports

0.34

 Excess

0.34

principalTable

0.32

Die

0.32

 Erfolgs

0.32

Activations Density 0.018%

numeric values or monetary amounts

lists ending with "and"

No Comments

No Known Activations

numeric values or monetary amounts

lists ending with "and"

No Comments

No Known Activations