INDEX

Explanations

phrases related to changes and calculations involving data or categorization

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

calculating, listing, comparing

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

endphp

-0.82

RegistryLite

-0.75

 esternos

-0.71

 OMITBAD

-0.68

 estekak

-0.67

ſammen

-0.66

tvguidetime

-0.66

 transférez

-0.66

 CURIAM

-0.66

 Normdatei

-0.65

POSITIVE LOGITS

 also

0.40

 likewise

0.35

 также

0.35

 inoltre

0.34

 bailarina

0.34

 również

0.33

 olje

0.32

 panneaux

0.32

 pintor

0.31

Activations Density 0.214%

phrases related to changes and calculations involving data or categorization

calculating, listing, comparing

No Comments

No Known Activations