INDEX

Explanations

references to various subjects or themes in documents

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

End of turn markers

np_acts-logits-general · gemini-2.0-flash

prior acts

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 صوتيه

-0.82

 Chwiliwch

-0.71

bewerken

-0.69

 surla

-0.68

AndEndTag

-0.67

 ComVisible

-0.66

 resourceCulture

-0.63

 esternos

-0.61

 Numerade

-0.60

Vidite

-0.56

POSITIVE LOGITS

 unrelated

0.33

 sexism

0.32

background

0.30

<?

0.30

 digress

0.29

 explot

0.29

borderTop

0.28

 salpic

0.28

Fall

0.27

erschutz

0.27

Activations Density 0.386%

references to various subjects or themes in documents

End of turn markers

prior acts

No Comments

No Known Activations