INDEX

Explanations

references to unseen or unnoticed aspects in various contexts

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

invisible, hidden, and transparent

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

 estekak

-0.46

 recrutement

-0.39

AccessorTable

-0.38

 orgánico

-0.38

 professora

-0.36

 désert

-0.35

 baterías

-0.34

 exclusivas

-0.34

 trône

-0.33

 espirituales

-0.33

POSITIVE LOGITS

 gyhoeddwyd

0.56

styleType

0.53

 invisible

0.49

 doInBackground

0.48

+#+

0.46

 silently

0.44

 autorytatywna

0.44

Попис

0.44

 Lightboxes

0.44

 invis

0.43

Activations Density 0.034%

references to unseen or unnoticed aspects in various contexts

invisible, hidden, and transparent

No Comments

No Known Activations